Project 1

1>

7 Standard parameters:

ECG, RESP, NIBP, SPO2, 2-TEMP,IBP, PR/HR. and ETCO2

Technical Specification of Patient Monitor 7 Parameter

Cardiac Output Module

• Width: 1.4 in ( 36mm)
• Height: 4.1 in (103.5 mm)
• Depth: 4.3 in (110 mm)
• Weight: 225 g ( 7.9 oz)

Spirometry Module

• Width: 1.4 in ( 36mm)
• Height: 4.1 in (103.5 mm)
• Depth: 4.3 in (110 mm)
• Weight: 250 g ( 8.8 oz)

SpO2 Module

• Width: 1.4 in ( 36mm)
• Height: 4.1 in (103.5 mm)
• Depth: 4.3 in (110 mm)
• Weight: 200 g ( <8.8 oz)

Temperature Range

• Operating: 0 – 40° C (32 – 100° F). When charging battery, or M3002A is mounted on back, or with Smart Hopping Interface 0 – 35°C (32 – 95°F)
• Storage: -20 – 60° C (-4 – 140° F)

Humidity Range

• Operating: 15% to 95% Relative Humidity (RH) (non condensing)
• Storage: 5% – 95% Relative Humidity (RH)

Altitude Range

• Operating: -500 – 3000 m (10000 ft)
• Storage: -500 – 4600 m (15000 ft)

Power Specifications

• Power Consumption: <70 W average
• Line Voltage: 100 – 240 V
• Current: 1.2 – 0.5 A
• Frequency: 50/60 Hz

Display Screen

• Size: 15 inch
• Type: 390 mm active matrix color LCD (TFT)
• Resolution: 1280 x 768
• Refresh rate: 59.9 Hz
• Useful screen: 334.1 x200.5 mm
• Pixel Pitch: 0.261 x 0.261

Network

• Standard: 10BASE-T and 100Base-TX (IEEE 802.3), auto-negotiation, full and half-duplex
• Connector: RJ45 (8 pin)
• Isolation: basic insulation (reference voltage: 250 V; test voltage: 1500 V)

USB Interface

• Standard: USB 2.0 high-speed
• Connector: USB series “Standard A” receptacle
• Power: Low power port 4.4 V min., max. load for all ports together 500 mA
• Isolation: none

Video Interface

• Connector: DVI-I (digital and analog, single link).
• Digital Video Connector: single link TMDS
• Analog video signals: 0.7 Vpp@75Ω
• HSYNC/VSYNC signals: TTL

Battery Operating Time

• New and fully loaded battery
  • With basic monitoring configuration: 2.5 hours (brightness set to optimum, MMS connected, NBP measurement every 15 minutes)
  • With extended monitoring configuration: 2 hours (brightness set to optimum, MMS and MMS extension connected, NBP every 15 minutes, Recorder, Pressure, Temperature modules connected)

Battery Charge Time

• When monitor is switched off: 3 hours
• When monitor is in use: up to 5 hours, depending on monitor configuration

ECG:

An electrocardiogram (ECG or EKG) records the electrical signal from the heart to check for different heart conditions. Electrodes are placed on the chest to record the heart's electrical signals, which cause the heart to beat. The signals are shown as waves on an attached computer monitor or printer.

If the test is normal, it should show that your heart is beating at an even rate of 60 to 100 beats per minute. Many different heart conditions can show up on an ECG, including a fast, slow, or abnormal heart rhythm, a heart defect, coronary artery disease, heart valve disease, or an enlarged heart.

RESP rate:

By Mayo Clinic Staff. Your respiratory rate is also known as your breathing rate. This is the number of breaths you take per minute. You can measure your breathing rate by counting the number of breaths you take over the course of one minute while you're at rest.

Normal respiration rates for an adult person at rest range from 12 to 16 breaths per minute.

NIBP:

An evaluation of non-invasive blood pressure (NIBP) monitoring on the wrist: comparison with upper arm NIBP measurement.

Non-invasive blood pressure (NIBP) measurement uses a pressurized cuff around the arm or the leg. The auscultatory method measures BP by detecting the sound of the opening and closing of the limb artery. The oscillometric method catches the vibration of the cuff during pressurized air release.

Healthy and unhealthy blood pressure ranges ; 120 – 129, and, LESS THAN 80 ; 130 – 139, or, 80 – 89. depends on the age factor.

SPO2:

Oxygen saturation (SpO2) is a measurement of how much oxygen your blood is carrying as a percentage of the maximum it could carry. For a healthy individual, the normal SpO2 should be between 96% to 99%. High altitudes and other factors may affect what is considered normal for a given individual.

A pulse oximeter can measure oxygen saturation. It is a noninvasive device placed over a person's finger. It measures light wavelengths to determine the ratio of the current levels of oxygenated hemoglobin to deoxygenated hemoglobin. The use of pulse oximetry has become a standard of care in medicine.

it can be measured with help of IR light and Red light pass through finger to photodectector.

TEMPERATURE:

For a typical adult, body temperature can be anywhere from 97 F to 99 F. Babies and children have a little higher range: 97.9 F to 100.4 F.

Mouth: Place the probe under the tongue and close the mouth. Breathe through the nose. Use the lips to hold the thermometer tightly in place. Leave the thermometer in the mouth for 3 minutes or until the device beeps.

in monitors the probe is placed on the skin.

Infrared (IR) temperature sensors enable accurate non-contact temperature measurement in medical applications. The most common applications for this type of temperature sensor is measuring ear temperature, forehead temperature, or skin temperatur.

The thermister or thermo couple is used in this case.

IBP:

Invasive (intra-arterial) blood pressure (IBP) monitoring is a commonly used technique in the Intensive Care Unit (ICU) and is also often used in the operating theatre. This technique involves direct measurement of arterial pressure by inserting a cannula needle in a suitable artery.

The most accurate method of obtaining blood pressure measurements is with the use of an invasive probe that is inserted directly into the lumen of an artery. An advantage of invasive monitoring is the ability to display blood pressure variations with each heartbeat.

At birth, a typical blood pressure is 80/50 mmHg. It rises steadily throughout childhood, so that in a young adult it might be 120/80 mmHg. As we get older, blood pressure continues to rise and a rule of thumb is that normal systolic pressure is age in years + 100.

The sensor, mainly used in the oscillometric method, is a pressure sensor (MEMS pressure sensor) made by the semiconductor process, which measures the pressure in the cuff using the air as a medium for pressure transmission.

PULSE RATE:

Your pulse rate, also known as your heart rate, is the number of times your heart beats per minute. A normal resting heart rate should be between 60 to 100 beats per minute, but it can vary from minute to minute.

Your age and general health can also affect your pulse rate, so it’s important to remember that a ‘normal’ pulse can vary from person to person.

To check your pulse at your wrist, place two fingers between the bone and the tendon over your radial artery — which is located on the thumb side of your wrist. When you feel your pulse, count the number of beats in 15 seconds. Multiply this number by four to calculate your beats per minute.

in this cane with healp of spo2 probe.

HEART ARTE:

A normal resting heart rate for adults ranges from 60 to 100 beats per minute. Generally, a lower heart rate at rest implies more efficient heart function and better cardiovascular fitness. For example, a well-trained athlete might have a normal resting heart rate closer to 40 beats per minute.

and it depends on the age factor and what work the person is doing.

in this case with help of ecg we can find out the heart rate.

ETCO2:

End-tidal carbon dioxide (ETCO₂) is the level of carbon dioxide that is released at the end of an exhaled breath. ETCO₂ levels reflect the adequacy with which carbon dioxide (CO₂) is carried in the blood back to the lungs and exhaled. Available evidence has established that ETCO₂ measurement can provide an indication of cardiac output and pulmonary blood flow. Non-invasive methods for ETCO₂ measurement include capnometry and capnography. Capnometry provides a numerical value for ETCO₂. In contrast, capnography delivers a more comprehensive measurement that is displayed in both graphical (waveform) and numerical form. For this reason, capnography is currently the most widely recommended method for monitoring ETCO₂

Non-invasive methods for ETCO₂ measurement include capnometry and capnography. Capnometry provides a numerical value for ETCO₂. In contrast, capnography delivers a more comprehensive measurement that is displayed in both graphical (waveform) and numerical form.

Sensors and transducer used:

A multi-parameter sensor patch can monitor the heart activity, body temperature, and blood oxygen saturation to provide a holistic view of the well-being of the person. The patch has multiple sensors-- ECG electrodes, temperature sensor, and optical sensors-- placed on the body.

he sensor consists of two parts: a light source and a photoelectric transducer. It is clamped on the fingertip or auricle of the patient.

2> the detail block diagram is shown in figure-1.

3>

The use of microcomputers in bedside monitors has revolutionized the acquisition, display, and processing of physiological data. There are virtually no bedside monitors or ventilators marketed today that do not use at least one microcomputer. Figure-2 shows a block diagram of a bedside monitor. Physiological signals such as the ECG are derived from sensors that convert biological signals (such as pressure, flow, or mechanical movement) into electrical signals. In mode computerized monitors, these signalsare digitized as close to the patient as possible.

Block diagram of a modern Bedside Monitor. Physiological signals from the patient are acquired by transducers. These transducers convert the appropriate physiological signal into an electrical signal that is then amplified and conditioned (usually an analog filter of some sort) and then present the signal to an Analog to Digitial converter (ADC) . The ADC sends the data to a microprocessor based signal processor which extracts features such as heart rate and blood pressure. After processing, the physiological signals are displayed on a display device and usually sent to a centralized ICU display system and frequently to a electronic patient record.

Some biological signals are already in electrical form, such as the currents that traverse the heart and are recorded as the ECG. The ECG voltage signal derived from the electrodes at the body surface is small—only a few millivolts in amplitude. The patient is electrically isolated from the bedside monitor, and the analog ECG signal is amplified to a level sufficient for conversion to digital data using an analog-to-digital converter (ADC). Digital data then can be processed and the results displayed (Weinfurt, 1990, p. 130) (Figure 17.4). As discussed in Chapter 5, the sampling rate is an important factor that affects the correspondence between an analog signal and that signal’s digital representation. Figure 17.5 shows an ECG that has been sampled at four different rates. At a rate of 500 measurements per second (Figure 17.5a), the digitized representation of the ECG looks like an analog recording of the ECG. All the features of the ECG, including the shape of the P wave (atrial depolarization), the amplitude of the QRS complex (ventricular depolarization), and the shape of the T wave (ventricular recovery), are reproduced faithfully. When the sampling rate is decreased to 100 measurements per second, however, the amplitude and shape of the QRS complex begin to be distorted. When only 50 observations per second are recorded, the QRS complex is grossly distorted, and the other features also begin to distort. At a recording rate of only 25 measurements per second, gross signal distortion occurs, and even estimating heart rate by measuring intervals from R to R is problematic.

Electrocardiogram (first and second traces), arterial pressure (third trace), and pulmonary-artery pressure (fourth trace) recorded from a patient’s bedside. Annotated on the recording are the bed number (E702), date (8 Jul 1989), and time (9:17:25). Also noted are a regular rhythm, a heart rate from the ECG (V) of 96 beats per minute, a systolic arterial pressure of 121, a diastolic pressure of 60, a mean pressure of 88 mm Hg, and a heart rate from pressure (PR) of 96. The patient is having premature ventricular contractions (PVCs) at a rate of three per minute; two PVCs can be seen in this tracing (at the beginning and near the end). The pulmonary-artery pressure is 29/11, with a mean of 19 mm Hg, and the blood temperature is 37.44˚C. The self-contained monitoring system has determined the values and generated the calibrated graphical plot.

Today, bedside monitors contain multiple microcomputers, with much more computing power and memory than was available in the systems used by the computer monitoring pioneers. Bedside monitors with built-in microcomputers have the following advantages over their analog predecessors (Weinfurt, 1990): ● The digital computer’s ability to store patient waveform information such as the ECG permits sophisticated pattern recognition and physiological signal feature extraction. Modern microcomputer-based bedside monitors use multiple ECG channels and pattern recognition schemes to identify abnormal waveform patterns and then to classify ECG arrhythmias. ● Signal quality from multiple ECG leads can now be monitored and interference noise minimized. For example, the computer can watch for degradation of ECG skin–electrode contact resistance. If the contact is poor, the monitor can alert the nurse to change the specified problematic electrode. ● Physiological signals can be acquired more efficiently by converting them to digital form early in the processing cycle. The waveform processing (e.g., calibration and filtering, as described in Chapter 5) then can be done in the microcomputer. The same process simplifies the nurse’s task of setting up and operating the bedside monitor by eliminating the manual calibration step. ● Transmission of digitized physiological waveform signals is easier and more reliable. Digital transmission of data is inherently noise-free. As a result, newer monitoring systems allow health-care professionals to review a patient’s waveform displays and derived parameters, such as heart rate and blood pressure, at the bedside, at a central station in the ICU, or at home via modem on a laptop computer. Figure 17.6 is a closeup of the signals and values from a typical bedside patient monitor. ● Selected data can be retained easily if they are digitized. For example, ECG strips of interesting physiological sequences, such as periods of arrhythmias (Figure 17.7), can be stored in the bedside monitor for later review. Today’s monitors typically store all of the waveform data from multiple leads of ECG and blood pressure transducers for at least 24 hours and sometimes for even longer. ● Measured variables, such as heart rate and blood pressure, can be graphed over prolonged periods to aid with detection of life-threatening trends. ● Alarms from bedside monitors are now much “smarter”and raise fewer false alarms. In the past, analog alarm systems used only high–low threshold limits and were susceptible to signal artifacts (Gardner, 1997). Now, computer-based bedside monitors often can distinguish between artifacts and real alarm situations by using the information derived from one signal to verify that from another and can confidently alert physicians and nurses to real alarms. For example, heart rate can be derived from either the ECG or the arterial blood pressure. If both signals indicate dangerous tachycardia (fast heart rate), the system sounds an alarm. If the two signals do not agree, the monitor can notify the health-care professional about a potential instrumentation or medical problem. The procedure is not unlike that performed by a human verifying possible problems by using redundant information from simpler bedside monitor alarms. Despite these. Figure-3 and Figure-4.

4>

1. Electrical safety.
2. isolation amplifier.
3. Defibrillation Protection.
4. EMI filter.
5. PASEMAKER PULSE ELEMINATION.
6. Earthing etc....

5>

Multi-parameter patient monitors (MPMs) have become increasingly important in providing quality healthcare to patients. It is well known in the medical community that there exists an intrinsic relationship between different vital parameters in a healthy person, these include heart rate, blood pressure, respiration rate and oxygen saturation. For example, an increase in blood pressure would lead to a decrease in the heart rate, and vice versa. Although it is likely to improve the performance of MPM systems, this fact is not explored in engineering research. In this work, experiments show that deriving additional features to capture the intrinsic relationship between the vital parameters, the alarm accuracy (sensitivity), no-alarm accuracy (specificity) and the overall performance of MPMs can be improved. The geometric mean of the product of all the vital parameters taken in pairs of two was used to capture the intrinsic relationship between the different parameters. An improvement of 10.55% for sensitivity, 0.32% for specificity and an overall performance improvement of 1.03% was obtained, compared to the baseline system using classification and regression tree with the four vital parameters.

Multi-parameter patient monitors (MPMs) are widely used in intensive care units (ICUs) and general wards to continuously monitor patients’ health based on the following human vital parameters: heart rate; blood pressure; respiration rate; and oxygen saturation (SPO2). MPMs in general require lower missing probability, probability for missing an alarm when an alarm is to be reported and false alarm probability, falsely reporting an alarm when there is no alarm to be reported. This means that the alarm accuracy, sensitivity and the no-alarm accuracy and specificity should be as high as possible.

Studies on physiological parameters show that there is a well-established intrinsic relationship between vital parameters in healthy people. For example, when a heart rate is on the higher side, arterial blood pressure is expected to be on the lower side, and vice versa, otherwise the health condition is said to be abnormal. However, to the best of our knowledge, MPMs do not take advantage of this relationship.

In this Letter, we capture the correlation (positive/negative) between human vital parameters, which include heart rate, arterial blood pressure, respiration rate and SPO2, to capture the intrinsic relationship between them and achieve higher sensitivity, specificity and overall classification accuracy. We use correlation features and geometric mean of the different vital parameters taken in pairs of two, in addition to the four vital parameters in the proposed system.

We use a classification and regression tree (CART), the decision tree learning algorithm using Gini-impurity index as the parameter for constructing the tree, to verify the effectiveness of the proposed approach in enhancing the performance of the MPMs. The system developed in the present work uses the MIMIC-II database to learn the trends in the vital signs.

2. Database used for the experiments

Experiments were performed with time series data such as heart rate, arterial blood pressure, respiration rate and SPO2, obtained from the MIMIC-II database. The MIMIC-II consists of two major components; namely clinical data and physiological waveforms. The physiological waveforms are collected from bedside patient monitors (Component Monitoring System Intellivue MP-70; Philips Healthcare) used in medical, surgical, coronal care and neonatal ICUs in a tertiary hospital. The waveform database includes high-resolution (125 Hz) waveforms (e.g. electrocardiograms), derived numeric time series such as heart rate, blood pressures and SPO2. The database also consists of monitor generated alarms along with the physician annotated files.

3. Decision tree algorithm for multi-parameter patient monitoring system

Decision tree learning is one of the machine learning approaches used for classifying data based on certain attributes. In this Letter, we train the decision tree with a continuous record of vital signs denoting normal and abnormal conditions of the patient. We identify the vital signs in the normal condition with the ‘no-alarm’ label and those in the abnormal condition with an ‘alarm’ label. The classifier is ideally trained to make the mapping x_i → y_i in a way that lowers the number of errors in the classification, where x_i are the input data points and y_i are the output labels. Once the system is trained with all possible cases of normal and abnormal conditions, it can then classify the vital signs for an unlabelled data set, to indicate when the patient's condition is deteriorating.

The system developed in this work uses the CART algorithm that learns the continuous multi-parameter patient data and the additional parameters to generate a tree model that makes decisions about the sample belonging to a normal or an abnormal patient condition. Decisions are made with the help of automatically generated rules. The CART algorithm is designed to produce a set of questions, the answers to which determine the next set of questions. The result of these questions is a tree structure, the ends of which are called terminal nodes, where the actual decisions are made. The algorithm makes use of the Gini-impurity index, a measure of node impurity; that is, it measures how often a randomly chosen element from the data set would be incorrectly classified if it were randomly labelled according to the distribution of the labels in the subset. Gini impurity can be computed by summing the probability of each item being chosen times the probability of a mistake in categorising that item. It reaches its minimum, zero, when all cases in the node fall into a single target category .The decision tree algorithm always attempts to minimise the deviance or impurity indicated by Gini impurity index. The Gini index of a target node t may be defined as:

IG=1−∑i=1Np(i/t)2

(1)

where N is the number of samples at the target node and p(i/t) is the relative frequency of class i at node t.

4. Improving the performance of MPMs

Towards improving the performance of the MPM without any additional algorithmic complexity, we use geometric means of the vital parameters taken in pairs of two, in addition to the four vital parameters used in the baseline system. The expanded feature set contains additional information about the balance/imbalance in the patient's health status: for example, the relationship between the parameters exposes the risk factor for problems like cardiovascular abnormalities, hypoxia (low oxygen content in the blood), hyperoxia (high oxygen content in the blood) and so on. In this Letter, we gather this additional information by increasing the feature set to F, thereby improving the system performance. The improved feature set F is

F=[x1,x2,x3,x4,x1x2−−−−√,x1x3−−−−√,x1x4−−−−√,x2x3−−−−√,x2x4−−−−√,x3x4−−−−√]

(2)

where x₁, x₂, x₃ and x₄ are the vital parameters, i.e. heart rate, blood pressure, respiration rate and SPO2 values, respectively, and the remaining six correlation features help capture the instrinsic relatioship between the vital parameters.

5. Experiments and results

Experiments were performed using the MIMIC-II database. The database consists of vital parameters from 413 patients. We first split the data into training and test data; training data consisted of 300 patients and the test data consisted of the remaining 113 patients. No patient data was shared between the training and test data. Subsequently, for computational considerations, we randomly selected 20 patients from the training data for training the decision tree algorithm and 8 patients from the test data set for the testing the effectiveness of the approach. The trials were then repeated ten times, every time selecting a new set of patients for the trial, and the results averaged across the trials. The training data at every trial consisted of more than 50 000 samples, and the test data consisted of more than 19 000 samples, each sample having an instance of all the four vital parameters. More than 90% of the samples represented no-alarm condition, and the remaining alarms.

In this experiment, we did not use any windowing techniques to preprocess the physiological signals, and therefore each sample was taken as an example to train and test the condition of the patient. Table 1 compares the performance results of the baseline decision tree MPM system using four vital parameters: heart rate; blood pressure; respiration rate; and SPO2, and the system with additional six correlation features. The total number of features in the baseline system is four, and the proposed system is ten, with an additional six correlation features.

There is a difference between medical device test equipment calibration and maintenance activities. Maintenance removes fault in a device, whereas calibration helps to detect the defective instrument that has failed to output values according to the manufacturer’s standard baseline readings.

The calibration activity varies from instrument to instrument. The manufacturer, in most cases, provides a built-in system for performing calibration processes. The manufacturer also recommends calibration frequency, although the user itself can devise its frequency, depending on the instrument application and frequency of use.

Every device has its tolerance limit, as defined by the manufacturer.

If the testing device’s calibration results fall within the tolerance limit, the device can be used further. On the other hand, if the calibration fails the device cannot be used for any process. Figure-5.

Corrective actions are required for devices that fail the calibration.

Such actions can either be performed in-house or the device can be sent to device specialists. Upon successful corrective actions, the device is again calibrated for verification. If results are satisfactory, the device is allowed to be used again. However, if the defect cannot be fixed, the instrument is permanently damaged and cannot be used further.

The calibration cycle is followed by complete documentation.

Common documentation includes:

• Calibration Plan, divided into monthly, quarterly, and yearly basis
• Calibration steps or a Standard Operating Procedure (SOP)
• Calibration Results Sheet indicating the actual results and tolerance limits

Calibration status is also indicated on the device, showing the calibration status, date of calibration, and next due date for calibration.

Equipment Calibration Related Regulations and Standards

The Medical Device Industry is strictly regulated by regulatory bodies such as Food and Drug Administration (FDA) in the United States to make them compliant with the regulations. Additionally, standardization bodies such as the International Organization for Standardization (ISO) also provide standards for performing medical device calibration.

Some common regulations and standards for medical device companies include the following.

• FDA 21 CFR Parts 11

The FDA 21 CFR Part 11 is the FDA’s guidance regarding electronically saved records. Electronic records act as an alternative to manual records. Compliance with this act makes the manufacturer’s electronic records trustworthy.

This regulation also guides electronic signatures, an alternative to manual signatures. However, this regulation does not apply to the electronic records transmitted by electronic means.

• FDA 21 CFR Parts 820

The FDA 21 CFR Part 820 is the FDA’s Quality System Regulation for production and process control of medical devices. This regulation allows manufacturers to provide valid inspection and test equipment results by establishing routine calibration procedures.

This regulation also bounds manufacturers to handle all medical devices so as not to harm their accuracy.

• Quality System Regulation (QSR)

Quality System Regulation (QSR) is the FDA’s quality policy for medical device manufacturers, similar to the Current Good Manufacturing Practices (cGMP) for Food, Drugs, and Biologics. The QSR only applies to finished medical devices that are ready to be used.

The QSR does not target any particular medical device but provides flexible guidelines to cover all medical devices. QSR puts the responsibility on the manufacturer to develop and establish their quality policies in light of QSR guidelines.

• ISO 13485:2016

ISO 13485 is a standard for medical device manufacturers developed by the International Organization for Standardization (ISO). It was reviewed and updated in the year 2016. The number 2016 comes with its terminology.

ISO 13485:2016 is the set of requirements for a Quality Management System (QMS) in the medical device industry.

This standard covers all life cycle stages of the medical device, including manufacturing, storage, distribution, installation, and servicing.

Recommended Reading: Medical Device Quality Management System (QMS)

6>

the gui interface is shown in figure-6.

7>

We report the establishment of the Multiparameter Intelligent Monitoring in Intensive Care II (MIMIC-II) research database that is notable for four factors: it is publicly and freely available to other research organizations upon request; it encompasses a diverse population of intensive care unit (ICU) patients; it contains high temporal resolution data, including laboratory results, electronic clinical documentation, and bedside monitor numeric trends and waveforms (such as the electrocardiogram); and it has been deidentified in a Health Insurance Portability and Accountability Act-compliant manner. The MIMIC-II database will support a diverse range of analytic studies spanning epidemiology, clinical decision-rule development, and electronic tool development.

Historically, large-scale ICU databases have been effective resources to understand risk factors and natural histories of critical illness as well as the efficacy of various treatment strategies. For instance, Acute Physiology and Chronic Health Evaluation I–III and Project Impact contained daily abstractions of patient data that provided new insights and scoring tools to relate patient outcomes and lengths of stays with the patients’ conditions on admission (1, 2). Such collection and analysis of large volumes of ICU data are invaluable to the advancement of clinical knowledge, but it is extremely effort-intensive because there are substantial challenges to the collection of the data. Such difficulties include: disparate sources of data, eg, clinical documentation versus laboratory results; erroneous or missing data; unsynchronized time references; proprietary data formats; limitations of computing power, networking bandwidth, and digital storage capacity; and concerns related to patient privacy. The challenge of data collection has sometimes been addressed through coordinated efforts by a network of clinical investigators interested in specific problem domains such as acute respiratory distress syndrome (ARDSNET Trial) (3), acute kidney injury (4), or septic shock (5). However, these powerful disease specific databases were not designed to be exploited as research resources to support other domains of ICU research nor are their data widely available.

In 2003, under National Institutes of Health funding, we established a research program with the objective of developing and evaluating advanced ICU monitoring and decision support systems. A critical requirement of our program was the development of a substantial and comprehensive clinical database from ICU patients. Now, 7 yrs later, the MIMIC-II database has reached a state of maturity sufficient to be made available to the wider research community. The database is intended to support a wide diversity of research in critical care. Unlike related databases, there are no access fees or extensive credentialing requirements, and documentation and other support are available so that the data will be accessible to the largest community of researchers.

This article contains a detailed report of the MIMIC-II data acquisition process, which was accomplished through collaboration among academic, industrial, and clinical groups. Summary statistics are provided to characterize the database and we provide examples of clinical hypotheses and physiologic signal processing algorithms we have studied with MIMIC-II. The high temporal resolution parameters within the database such as hourly vital sign trends, ventilator settings, intravenous medication drip rates, and fluid balances enable novel investigations of transient clinical outcomes such as hypotensive episodes. Similarly, MIMIC-II enables the analysis of transient independent variables such as electrocardiogram waveform features and their associated clinical outcomes. The unique features of MIMIC-II are compared with other major databases and we discuss the major challenges encountered in developing MIMIC-II and explore future improvements. The MIMIC-II database takes advantage of improvements in healthcare information technologies to establish a new standard in public-access databases for critical care research.

Go to:

MATERIALS AND METHODS

This study was approved by the Institutional Review Boards of Beth Israel Deaconess Medical Center (Boston, MA) and the Massachusetts Institute of Technology (Cambridge, MA). Requirement for individual patient consent was waived because the study did not impact clinical care and all protected health information was deidentified.

• Patient Population

This first release of the MIMIC-II database encompasses virtually all adult patients admitted to ICUs at Boston’s Beth Israel Deaconess Medical Center during the period 2001–2007; additional MIMIC-II data collection is ongoing. Boston’s Beth Israel Deaconess Medical Center is a 620-bed tertiary academic medical center in Boston and a level I trauma center with 77 critical care beds. The ICUs are closed with 24-hr inhouse intensivist supervision of patient care. These ICUs include the medical, surgical, coronary, and cardiac surgery recovery care units. ICU stays separated by >24 hrs were counted separately even if they occurred within the same hospital stay.

• Database Development

The data acquisition process was not visible to staff and did not interfere with the clinical care of patients or methods of monitoring. Two categories of data were collected: clinical data, which were aggregated from ICU information systems and hospital archives, and high-resolution physiological data (waveforms and time series of derived physiological measurements) that were obtained from bedside monitors.

• Clinical Data

Clinical data were obtained from the CareVue Clinical Information System (models M2331A and M1215A; Philips Health-care, Andover, MA) deployed in all the study ICUs as well as from hospital electronic archives (Table 1). The data included such items as time-stamped nurse-verified physiological measurements (eg, hourly documentation of heart rate, arterial blood pressure, pulmonary artery pressure, etc); nurses’ and respiratory therapists’ progress notes; continuous intravenous drip medications; fluid balances; patient demographics; interpretations of imaging studies; physician orders; discharge summaries; and International Classification of Diseases, 9th Revision (ICD-9) codes. Comprehensive diagnostic laboratory results (eg, blood chemistries, complete blood counts, arterial blood gases, microbiology results) were obtained from the patients’ entire hospital stay, including periods outside the ICU.

• Physiological Data

Physiological data were obtained with the technical assistance of the monitoring system vendor. Patient monitors (Component Monitoring System Intellivue MP-70; Philips Healthcare) were located by every ICU patient bed. Each monitor acquired and digitized multiparameter physiological data; processed the signals to derive time series (trends) of clinical measures such as heart rate, blood pressures, and oxygen saturation, etc; and also produced bedside monitor alarms. Those data were all transmitted to a networked nursing central station within each ICU (M3155 Intellivue Information Center; Philips Healthcare). The physiological waveforms (such as electrocardiogram, blood pressures, pulse plethysmograms, respirations) were sampled at 125 Hz, and trend data were updated each minute. The data were subsequently stored temporarily in a central database server that typically supported several ICUs. A customized archiving agent, developed through collaboration with Philips Health-care, created permanent copies of the physiological data residing in central database servers. The data were physically transported from the hospital to the laboratory every 2 to 4 wks where they were deidentified, converted to an open source data format (6), and incorporated into the MIMIC II waveform database. Limited capacity and intermittent failures of the archiving agents restricted waveform collection to a fraction (15%) of the monitored ICU beds (Table 2). No attempt was made to assure that the ICU records with waveform/trend data were statistically representative of the database as a whole.

• Database Merger and Postprocessing

The second stage in developing the MIMIC-II database involved significant data postprocessing and database organization to obtain integrated medical records for each patient. Across the hospital’s clinical databases, patients are identified by their unique Medical Record Numbers and their Encounter Numbers (the latter uniquely identifies a particular hospitalization for patients who might have been admitted multiple times), which we relied on in merging information from different hospital sources. Matching waveform records to clinical data was based on unique identifiers such as medical record numbers, dates of admission, and patient names. Sometimes, however, nurses did not enter patient identifiers into the bedside monitors and as a result, only approximately half of the available waveform records could be uniquely matched to clinical data. More information on database merger, in particular, how database integrity was ensured, is available at the MIMIC-II web site (http://physionet.org/mimic2).

An additional task was to convert the patient monitoring data from Philips’ proprietary format into an open-source format. With assistance from the medical equipment vendor, the waveforms, trends, and alarms were translated into WFDB, an open data format that is used for publicly available databases on the National Institutes of Health-sponsored PhysioNet web site.

Because MIMIC-II is intended to be a reflection of real-life clinical data (rather than pristine data derived from a carefully conducted clinical trial), the clinical and physiological content of the database has not been altered. In other words, we did not enforce any range restrictions or other sanity checks on the data (beyond what each individual hospital database might impose).

• Deidentification and Health Insurance Portability and Accountability Act Compliance

All data that were integrated into the MIMIC-II database were deidentified in compliance with Health Insurance Portability and Accountability Act standards to facilitate public access to MIMIC-II. Deletion of protected health information from structured data sources was straightforward (eg, database fields that provide the patient name, date of birth, etc). As well, we removed protected health information from the discharge summaries, diagnostic reports, and the approximately 700,000 free-text nursing and respiratory notes in MIMIC-II using an automated algorithm that has been shown to have superior performance in comparison to clinicians in detecting protected health information. This algorithm accommodates the broad spectrum of writing styles in our data set, including personal variations in syntax, abbreviations, and spelling. We have posted this algorithm in open-source form as a general tool to be used by others for deidentification of free-text notes.

• Database Distribution and Documentation

The MIMIC-II database was developed with the intention to make its contents freely available to interested researchers. The Internet is an effective distribution mechanism to facilitate the dissemination of the deidentified MIMIC-II database. To restrict traffic to legitimate medical researchers, access requires completion of a simple data use agreement and proof that the researcher has completed human subjects training.

The MIMIC-II database is available in two forms. In the first form, interested researchers can obtain a flat-file text version of the clinical database and the associated database schema that enables them to reconstruct the database using their method of choice. In the second form, interested researchers can gain access to the database through a password-protected web service. Database searches require the users to familiarize themselves with the database layouts and to program database queries using the Structured Query Language. Query output can be exported to comma-separated files to be analyzed offline using statistical or other software. Accessing and processing data from MIMIC-II is complex. It is highly recommended that studies based on the MIMIC-II database be conducted as collaborative efforts that include clinical, statistical, and relational database expertise.

Detailed documentation and procedures for obtaining access to MIMIC-II are available at the MIMIC-II web site (http://physionet.org/mimic2).

• Database Characterization

We report the characteristics of version 2.4 of the MIMIC-II database (released on February 1, 2010) so that investigators can determine whether a potential study of interest is possible with MIMIC-II. Clinical data that are summarized include mortality, length of stay in the ICU, primary ICD-9 codes, patient demographics, and frequency of use of significant therapeutic interventions. The source code used to generate the statistical results provided in this article is publicly available.

Acuity scores were not routinely documented for MIMIC-II patients during the admission process. We implemented an automated algorithm to retrospectively compute the ICU Simplified Acute Physiological Score (SAPS) I scores for the first 3 days of all admissions with complete SAPS I data. The SAPS I formula was chosen for its simplicity, requiring only available clinical laboratory measurements, fluid balance, and vital signs (There have been several refinements to the original SAPS algorithm that incorporate the presence or absence of chronic disease (such as cancer, AIDS, etc) in the overall acuity of a patient. Such clinical data exist in MIMIC-II free-text discharge summaries.). The mortality rate was trended as a function of admission SAPS I scores as a method to validate the automated SAPS I calculations. The distribution of problems, mortality rates, and acuity was also analyzed across the different adult ICUs (coronary care unit, medical ICU, surgical ICU, cardiac surgery recovery unit).

To illustrate some of the analyses possible with the database, we computed the incidence, associated mortality, and odds ratios of a range of diseases that can be defined by objective laboratory abnormalities and of vital sign abnormalities, including heart rate, systolic blood pressure, respiratory rate, and arterial oxygen saturation. We only included the population in whom those diagnostic data were ordered. We excluded the small population with multiple ICU stays in case their characteristics were unrepresentative, and we excluded the final 12 hrs of any ICU stay to avoid confounding the results with physiology associated with withdrawal of care.

Lastly, We compiled a comparison of MIMIC-II to other ICU research databases.

8>

The intensive care unit (ICU), designated as high risk area from infection prevention point of view, has a concentration of lifesaving medical aid and nursing care for patients who are critically ill. The engineering controls both during design, commissioning and operation play a major role. The designer of an ICU should not only have insight of clinical work flow and protocol but have a sound knowledge of engineering controls. The design consultant should capture requirement of patients, intensivists, surgeon, nurses, technicians and infection control officer and in-house engineers (both general and clinical) for efficient design of ICU. The fine balancing act of quality, cost and time factors during design planning and execution will yield better results to hospitals on a long-term basis. Any flaws in ICU engineering control may result in serious compromise in patient care as evident from above mentioned examples.

The following facility related points are required to be considered as part of facility planning and further engineering controls during operation.

Space programme

The guidelines for ICU design should be based on criteria set by ISCCM, India given below.
Level I, six to eight beds —small district hospital, small private nursing homes, rural centres.
Level II, six to eight beds — larger general hospital
Level III, 10 to 16 beds-tertiary level hospitals

The new level III ICU are further planned based on type/ usage of ICU i.e. general or speciality-based like medical, cardiac, neurosurgical, transplant, paediatric. The location of ICU should be close to the operation theatre, imaging diagnostic services and laboratory.

The floor space area size of each patient bed space for ICU can be decided based on classification of open space ICU or cubicle/ separate room. The floor spaced per ICU bed can be planned 3 m X 4 m (12 sq. m) to maximum of 5 m X 5 m (25 sq. m) based on consideration of application of ICU including services and equipment positioning. The outside environment viewing window for each patient bed is strongly suggested as part of the design. A minimum of one to two metre distance should be kept between two beds as per NABH standard.

The height between floor and false ceiling should be three metres to facilitate to bring in engineering services from the ceiling. The ideal single leaf door size for each ICU cubicle or separate room having clear space of 2.1 m X 1.2 m height and width with wide view panel for visual access to patient is suggestive. It is indicative to have 12-16 beds per ICU area for optimal design considering all essential support functions. The total area of ICU should be 2.5 to three times the total space of ICU beds which includes supply and service corridor/ passageway of 2.4 metre width. At least one patient cubicle as isolation with anteroom facility having negative pressure is recommended within the ICU.

The overall design of the ICU should consider patients, staff and visitor movement, storage space of equipment and medicine, location of essential areas like nurse station, clean and dirty utility etc. The other essential areas in floor plan of ICU may contain nourishment room, stat lab, linen storage, staff lounge and utility services.

8 multiparameter monitor is connected to the partients and measure the parameters and all together is combined in the form of wifi network and connected to one nursing station that will reduce the effort of nurses to monitor all the patients vitals.

The healthcare industry is strongly acting towards addressing healthcare professional’s pain points. For an instance, the central patient monitoring systems addresses once such pain point of monitoring a large set of patients. The central patient monitoring systems are computer systems that are networked with a fixed number of patient monitors.

The central patient monitoring system displays the respectively connected monitoring device waveforms on a single or multiple monitors. This allows the healthcare professionals to monitor and keep a check on the number of patients at the same time. Most of the central patient monitoring system allows the monitoring of at least two waveforms. Depending on the configuration, the central patient monitoring system may include modules for various parameters such as ECG, respiratory rate, body temperature, SpO2, cardiac output and others.

The receivers of the central patient monitoring systems are connected to the bedside monitors of each patient. Some telemetry devices also allow the central patent monitoring system receiver to be connected. The main monitor of the central patient monitoring system is then placed in the central station. The central patient monitoring systems are used in the general medical area and surgical areas, cardiac rehab centers, they can also be used with telemetry devices.

A typical central monitoring system has a life time of about 7-10 and may cost US$ 4,500 - 40,000. The central patient monitoring system has only one major requirement, which is uninterrupted power supply. Furthermore, the central patient monitoring system also has the capacity to save the data for a limited period of time. This allows the healthcare professional to review the data when required.

Generally, central patient monitoring systems are able to network with a limited number of patient monitoring systems. For an instance, the central patient monitoring system can be connected to 8, 16, 32, 64, 72 and sometimes even up to 240 patient monitors. Most of the central patient monitoring systems are able to identify most kinds of abnormal waveform. When an abnormal waveform is detected the system is programed to trigger a code alarm. This simplifies the need to monitor the central patient monitoring system continuously and closely.

However, the use of central patient monitoring systems is still debatable since it is believed that the hospital personals may pay more attention to the equipment more than the patient itself. Many physicians believe that even a functioning reliable central patient monitoring system cannot replace the frequent and direct observations. Also, there is always a risk of false alarms that can alarm fatigue as a result the staff may miss an actually critical patient event. These are some of the factors that may hinder the growth of the central patient monitoring system market.

how the all monitors are connected to central monitor is shown in figure-7 And figure-8.

Remote patient monitoring systems are designed to obtain a number of physiological data from patients. Most common data are Electrocardiogram (ECG), Electroencephalogram (EEG), heart beats and respiration rate, oxygen volume in blood or pulse oximetry, signals from the nervous system, blood pressure, body/skin temperature and blood glucose level. In addition to these, sometimes, weight of the patient, level of activity of the patient and sleep data are collected. A number of researches have been done for wound management and sleep monitoring applications.

Traditional systems collect data using sensors attached to the body. But these systems pose difficulty in terms of mobility for the patient and a patient’s activities in daily living. Since this kind of devices influence patient’s comfort, sensitive physiological data get influenced. So the readings may not represent a patient’s actual illness but rather the discomfort a patient, especially an elderly patient, is undergoing at the time a physiological reading is being taken. While there are many situations that invasive or with -contact interventions are necessary, novel research look into methods for obtaining physiological data as non-invasively as possible. For this end, contactless methods are being researched on extensively during the past few years. In this paper we discuss both with-contact (i.e. contact-based) and contactless methods as both these types of systems are important for a researcher in the field.

The remainder of this section briefly lists past general reviews on remote health monitoring systems. Disease-specific reviews are given in the Sect. 3 under each disease type.

General surveys on contact-based RPM systems Remote patient monitoring concept first came up with contact-based methods. These systems use different sensors, processing technologies, communication technologies, post-processing actions, databases and receivers/end-terminals. They also target different diseases. The variety in systems may exist due to the designer’s choice or the desired outcome of other special features (embedded, low cost, low power etc.) the system designer planned for its users. For example, the designer may have been targeting a system with increased power efficiency or medication management.

This section gives an introduction on recent survey papers on contact-based remote monitoring systems.

In the review paper (Silva et al. 2015), developments in Mobile Health (m-Health) and mobile health applications in the market are discussed. This paper reviews papers on applications such as foetal heart rate monitor using mobile phone, medical health record system for paediatric human immunodeficiency virus (HIV), haemoglobin meter, mobile phone image transmission for diagnosis in addition to typical applications such as pulse oximeter. Reviews on two studies which use serious gaming approaches for learning ultrasound needle placement skills and collaborative exergaming applications for child obesity epidemic using peer-to-peer architectures have been given (Silva et al. 2015). Radio Frequency Identification (RFID) positioning, bio-patch and system-on-chip systems, cognitive stimulation through digital TV, ViCare (Hossain and Ahmed 2012)—an elderly people monitoring and virtual reality based surgical simulator are some significant work discussed in this paper. Additionally, it reviews a few papers based on technological improvements of WBSN and Wi-Max and ICT platforms generally.

In survey paper (Sawand et al. 2015) security issues in wireless body area networks and some existing solutions are discussed. It describes eHealth systems to have three key usages such as the patients, medical service providers and health professionals. They present architecture of eHealth systems with layers in terms of functionality. The paper also stresses on the importance of different security aspects such as patient’s information, physiological data recordings and about the patient’s environment. In a remote monitoring system, these data will be obtained by sensors and the sensor system together with other system components will need to have advanced security and privacy protection features in order to face any malicious attacks. Any malicious attack can harm the software as well as the functionality of the hardware which will pose a greater threat to the patient. Also emphasis is given on ensuring safe device-to-device communication and further data transmission (beyond the WBAN). Some important remedies for these issues have been suggested as role based access control, use of data based on accountability and revocability in case of a malicious attack. For the security in the communication and network level, secure routing protocols, cipher based algorithms and digital signature usage are presented. For the transmission of information beyond the personal area network, the paper suggests data encryption, data integrity, freshness protection and authentication. In Sawand et al. (2015) some existing solutions are given for data storage related security breaches and data access related security breaches. For the former case, they show the findings of another research where data is encrypted and divided into a number of blocks which will be saved in a number of different nodes. The data blocks will have algebraic signatures which will be useful for identifying the data. This method has been understood as to have low computational overhead. For the data access related security breaches, good countermeasures shown are symmetric key cryptography based schemes and public key cryptography based schemes. Biometric cryptography using key management for the inter-WBAN communications are also presented. Use of advanced image and signal processing is also suggested.

Another important aspect of this paper is to discuss energy management in WBANs. Power-saving hardware and light weight routing algorithms are suggested. Usage of hybrid encryption techniques like elliptic curve cryptography has been cited in Francis et al. (2015). This further speaks about weighted user access control, other encryption methods and data encryption such as set partitioning in hierarchical trees (SPIHT) based encryption method used for encrypting ECG signals. Further, a security architecture for eHealth systems is presented in this paper.

The survey paper (Chen et al. 2014) is on advancement of machine to machine communications in healthcare, energy management and entertainment applications. According to the paper, machine to machine communication systems are composed of three domains; namely, machine to machine device domain, network domain and application domain. The device domain is where the data collection, processing and issues like power saving of the devices will be considered. In the network domain, making compatible network between large co-existing points are addressed while in the application domain various monitoring, controlling and management services are focused. Then, the remaining discussion of the paper has been focused to wireless body area subnetworks and use of mobile cloud computing in the area of remote healthcare. Some of the commercially available sensors are given here such as accelerometer/gyroscope, wireless pulse oximeter sensor, electromyogram (EMG) sensor, electroencephalogram (EEG) sensor, various environmental status monitoring sensors (such as carbondioxide-CO₂ sensor, humidity, temperature sensor), infrared (IR) sensors, occupancy and motion detection sensors, and power sensors. Most importantly, various important issues in machine to machine communications; interference, channel dynamics, resource constraints, device heterogeneity, self-organization, quality of service (QoS) Support and Security are also raised in the paper.

Security issues in WBANs, ‘the necessity of secure and dependable distributed data storage, and fine grained distributed data access control’ are discussed in the review paper (Mainanwal et al. 2015). It also discusses the importance of the following: data confidentiality, data integrity, data authentication, data freshness, secure localization, availability and secure management.

Another recent brief survey (Tsakalakis and Bourbakis 2014) compares some existing systems. Recent work based on wearable sensor based systems, implantable sensor systems and portable and implantable biosensor based systems are compared.

A lot of important developments in the field such as cuff-less blood pressure meter, wrist-watch based fall detection method and cane-based fall detection system are reviewed in Patel et al. (2012). Some existing systems for infant monitoring are analysed in Zhu et al. (2015). In the paper (Hu et al. 2013), discussions about photoplethysmography technology and some comparisons between these systems with existing with-contact methods are done.

General surveys on contactless RPM systems Contactless remote health monitoring methods can be categorised mainly into two sections; image-based methods and radar-based methods. Image-based systems analyse images of patients as their data and detect illnesses or falls, radar based methods used radio frequencies to get inputs for their systems and also sometimes have patient localization capability. These systems are comparatively new (only few years since inception) so there are a lot of areas for improvement. Although it is the case, some works of research have achieved good outcomes. The remainder of this section discusses some of the very recent reviews on contactless image-based patient monitoring methods.

McDuff et al. (2015) review photoplethysmographic imaging methods for non-contact remote health monitoring. This technique uses camera colour signals that are processed to recover the blood volume pulse signal out of which physiological data such as pulse rate, pulse rate variability, respiration rate and blood oxygen levels can be obtained.

A recent survey gives an insight to the studies carried out in vision based patient monitoring (Sathyanarayana et al. 2015). They present the analysis in seven different applications (including detection of illnesses, sleep monitoring and activity monitoring) and the advancement of vision-based patient monitoring technology for these applications. Use of single cameras, multi-camera arrangements, infrared cameras, time-of-flight cameras and other sensors such as bio-inspired vision sensors and monocular vision sensors is discussed.

Baig and Gholamhosseini (2013) compare several smart health monitoring systems and discuss challenges and issues in the current systems. This paper analyses the advancements of health monitoring systems under two categories; smart systems and traditional systems. Smart systems include remote health monitoring systems, mobile health monitoring systems and wearable health monitoring systems while traditional category is on wired health monitoring systems (i.e. general health monitoring systems).

Erden et al. (2016) has done a very recent survey on sensors used in assisted living. In this review, the focus is on signal and image processing methods used in recent systems. They point out that assisted living systems should be low cost, with high accuracy and user acceptance while there should also be mechanisms for better privacy and protection of data while sharing the minimal private data essential for being monitored remotely. The review also stresses that machine-learning algorithms customized for sensors should be used as there are a number of various sensors (passive infrared sensors, vibrational and acoustic sensors, cameras, etc.) used in a system for assisted living. Several algorithms are used for these multi-sensor systems. These algorithms include Discrete Fourier Transform, Discrete Wavelet Transform, complex wavelet transform techniques etc. Classifiers mentioned in the review include machine learning methods such as support vector machines and k-nearest neighbour methods.

With the above reviews available in the literature, the authors of this paper review the studies during the period 2012–2016. A few very recent studies (from 2017) have also been included considering their importance for a reader who would pursue further research in this area. As mentioned before, the systems are categorized based on the targeted disease. We chose only four categories based on the criticality and prevalence of them in the present society. As such, the scope of this paper is limited to the following four categories which will be discussed in detail in Sect. 3.

1.

Cardiovascular System and Respiratory System related diseases

2.

Fall detection and Mobility related diseases

3.

Brain, Neurological Disorders and Mental Health

4.

Diabetes

The above four sections were decided to be included in the view that most of the patients monitoring technologies target the ailments that fall under the above four. In each category, two further classifications present contact-based and contactless efforts.

HOW inter-WBAN communications works.

We assume a WBAN to consist of sensor devices that are capable of measuring biometrics related to human body and also a high power and high storage device known as personal server (PS), which can be a laptop or a hand-held device. Medical server (MS) receives all the information collected by PS through the sensor nodes. All sensor nodes are directly connected to their relevant PSs. Sensor nodes measure biometrics and forward them to the PS. PS in turn transmits collected information to the MS through the internet. Each WBAN is associated with one body. Multiple WBANs are associated with the central MS. PS can communicate with other PSs as well as the MS. The MS stores and processes the information of all the WBANs that are associated with it.

All sensor nodes are constrained in energy because they use rechargeable batteries. Sensor nodes are ordinary devices with limited computation, communication, energy supply, and storage capabilities. PS is a powerful node and has more computation, communication, energy supply, and storage capabilities. We assume that the PS is preloaded with node identities and relevant keys before deployment. Keeping in view the storage constraints, in intra-WBAN communication, only one key is preloaded in sensor nodes before deployment. The system architecture of the WBAN, as per our assumptions, is shown in Figure 9. The application scenario of inter-WBAN communication includes multiple bodies under surveillance and all bodies communicate to a remote base station, like in the battlefield, the soldiers are deployed in the enemy territory and they communicate to the remote base station in their own territory. As in Figure 9, the PSs of all the bodies communicate to the base station and then through Internet to the remote MS.

• Proposed Scheme for Intra-WBAN Communication

Intra-WBAN communication includes the communication of sensor nodes with the PS. We propose a hybrid approach for key management in intra-WBAN communication. Due to the memory limitations, only a single key named as secret key KSN,MSKSN,MS is preloaded in the sensor nodes and is used in case of PS compromise. Other keys are generated by sensors themselves using their biometrics. The process is done in two steps: feature generation and key agreement.

Feature Generation

In the feature generation phase, features are extracted and then quantized for secure intersensor communication with the help of EKG using discrete wavelet transform (DWT). DWT allows good localization both in time and spatial frequency domains and is computationally inexpensive. In the process of communication between SNs and PS, sensors sample the EKG signal at the sampling rate of 125 Hz in time duration of 5 seconds. To remove unnecessary frequency components, the signal is then filtered. 625 samples are produced by five-second sample of EKG and then divided into 5 parts of 125 samples each. DWT is applied on each part after applying filtration. The 320 coefficients feature vector is formed by concatenating the 64 coefficients horizontally. In the quantization phase, the generated feature vector is divided into 20 blocks, each containing 16 coefficients, and then they are quantized into a binary streaKey Agreement

After the process of quantization, creation of feature vectors, and formation of blocks, the key agreement process is done. In the key agreement phase, PS broadcasts data request message as shown in message m1m1 of Figure 10, consisting of ID_PS, DataReq, and nonce. All sensor nodes which have the required data first compute the shared pairwise key with the PS by applying keyed hash function on feature blocks, ID_PS and ID_SN as follows:

KPS,SN=HMAC((b11,N)⋯(b211,N),IDPS∥IDSN)m1:PS⟶*:IDPS,DataReq,noncem2:SN⟶PS:IDSN,EKPS,SN(IDSN,Data), MACKPS,SN(IDSN,nonce,Data). FIgure-10.

SN encrypts the data with key KPS,SNKPS,SN and also computes MAC on ID_SN, nonce, and data using the same key KPS,SNKPS,SN. SN sends its ID, encrypted data, and MAC to the PS as shown in message m2m2. When PS receives this message, first it calculates the KPS,SNKPS,SN by applying the keyed hash function on the feature blocks, ID_PS and ID_SN. As feature blocks are the same on both sides, KPS,SNKPS,SN generated by PS will be same as that of SN. Incorporation of ID_SN in key generation process ensures the establishment of unique pairwise key of PS with all communicating SNs. PS decrypts the message with KPS,SNKPS,SN and compares ID_SN and received feature blocks (data) with decrypted message ID_SN and feature block on PS to ensure that both parties have generated the same key. The message authenticity is checked by PS through MAC verification with KPS,SNKPS,SN.

In Figure 10, the key agreement phase is shown. EKG signal is used for feature generation. Feature vector of 320 coefficients is generated by concatenating 64 coefficients horizontally. These generated features are then divided into 20 blocks of 16 coefficients and then quantized into the binary stream. PS sends data request with its ID to SNs. SNs generate the shared pairwise key with PS by applying keyed hashing on the feature blocks and IDs of both PS and SN. The resulting key is used to encrypt data requested by the PS . Data is verified through MAC verification of the PS. Hamming distance is calculated to verify that the data blocks of both the sender and the receiver are the same.

10>

Forn any parameter exceeding or below the normal range its giving sound alaram and indication of some colour lites.

1. excess of normal range it will give red light and beep sound
2. below the normal range its giving yellow light and beep sound
3. for normal range greenlight no beep sound.

Alarms are typically displayed in two ways or as a combination of both:

1.

Acoustic

The alarm is given as a warning sound. Most manufacturers distinguish the priority of an alarm with different signals. Intuitive alarms with different tone sequences (e. g. 'short-long-short' for 'ven-ti-late') have been the object of research but have not found their way into routine clinical practice. Alarms directly mentioning organ systems, device hardware, or parts of it (e. g., ventilation or circulation) or alarms with direct labeling of the physiological problem ('blood pressure' or 'oxygen') have also not been introduced into practice.

2.

Visual

Visual alarms involve mostly flashing or coloring of the related parameter in an eye-catching manner. Some systems provide integrated displays of several parameters. One example is a spider-display, which shows the relationship of different parameters in a stylized spider web. Such applications can be useful to display different parameters in context. Compared to other professions in industry and aviation, adoption of such new displays in healthcare has been slow.

11>

USB is an industry standard, ‘user-friendly’ method of transferring data between a host device (such as a computer) and a peripheral device (for example, a mouse). To most computer users, the system simply allows the use of various devices by attaching them via a USB port.

USB stands for Universal Serial Bus. A ‘bus’ within a PC is a collection of wires that transfer data between components inside the computer, or between the computer and its peripheral devices, much as an electronic busbar distributes power throughout certain large, power-hungry environments such as factories and data centres.

Before the launch of USB, each peripheral device was attached to a computer with its own individually shaped port. As the number of peripheral devices increased over the years, a new standardised means of transferring data between the main host and a range of devices was sought. This ultimately resulted in the development of USB.

When a peripheral device is attached via USB, the host computer will detect what kind of device it is and automatically load a driver that allows the device to function.

Data is transferred between the two devices in small amounts known as ‘packets’. A set number of bytes (a unit of digital information) is transmitted with each packet.

Other information is also sent, including:

• the source of the data
• the destination of the data
• the length of the data
• details of any errors that have been detected

There are four types of data transfer that can occur:

Interrupt transfer. Peripheral devices such as keyboards and mice use this type of data message to send smaller amounts of data. Such transfers are often used for less frequent but important requests. The devices generate the requests, though they must wait for the host to inquire about the specific data the remote device needs.

Such requests are guaranteed to be reattempted if the first transfer fails. These transfers will also let you know about any changes to the status of the device.

Bulk transfer. Used by printers and digital scanners for large amounts of data, this type of transfer is low-priority and not time-critical. The transfer will slow down if the host computer has a number of USB devices connected.

Isochronous transfer. Audio, video and other real-time data uses isochronous transfer. Errors can occur during the transfer, though the transfer will not be interrupted in order to resend the packets. However, such transfers usually involve situations where the accuracy of the data is not critical, such as audio elements that may not be picked up by the listener. Missing these elements is preferable to retrying data, which could result in glitching audio.

Control transfer. This type of data transfer is used to configure and control a USB device. The host sends a request to the device and the data transfer follows. Control transfers are also used to check status. Only one control request is handled at any one time.

USB cables are capable of transferring both power and data. To achieve this, every USB cable features two sets of wires. One set carries the current while the other transfers the data signals.

Within the standard USB 2.0 connector you can see four metal strips. The outer two strips are the positive and ground of the power supply. The two central strips are dedicated to carrying data.

With the newer USB 3.0 connector, the data transfer speed is increased by the addition of extra data-carrying strips; four extra signalling wires help USB 3.0 achieve its super speed.

USB 2.0 transfers data at a top speed of 480 megabits per second (Mbps), while USB 3.0 can transfer data at up to 5 gigabits per second (Gbps).