Detection of medical instruments project- PART 2

HEALTHCARE
MEDICAL INSTRUMENT
DETECTION SYSTEM

With respect to the futuristic developments in image processing technology, it is expected
to gain market growth from 2020 to 2027. It is expected to reach USD 25,702 million by
2027 (growth rate of 21.8%). Healthcare is one of the prime industries which requires AI
and Image processing techniques to provide better classification and diagnostic results.
Latest data shows that:
Increase in the manufacturing of ventilators to treat COVID-19 patients.
Increased production of pulse oximeters and oxygen concentrators with reference to
COVID-19.
Several startups have used their platforms to promote portable and advance versions of
medical devices. (AgVa healthcare)
The conglomeration of A.I. and the human mind can transform
millions of lives. Its true potential lies in the welfare of living beings.

OBJECTIVES
 The prime objective is to identify and classify the given image among the 20 most
popular medical devices and provide a brief overview on the detected instrument.
 A dataset containing various images of biomedical devices or instruments. Trusted and
commercial sources are used to gather information about the present scenario.
 An algorithm for digital image processing. This can include any pre-trained neural
network.
 To test, train and evaluate the performance of the algorithm containing a lot of features.
 Utilizing this algorithm for classifying various images of biomedical instruments.
 A website, virtual portal or an app by which the input images are uploaded and
processed to get the required output.
 Along with the output, some information is provided to develop a basic understanding of
the medical device.

COMPONENTS USED
 Convolutional Neural Networks (CNN) are a specific application of deep learning in the field of
images. This network is used to classify or perform other operations on images. The primary
function of CNN’s is to extract features from images basically they convert multi-dimensional
images to 1 dimensional vectors. These CNN’s are combined with Fully Connected layers to
process the vector.
 VGG-16 is one such deep CNN which was developed by the Oxford University. In this project this
model has been used for the detection of medical devices.
 The training dataset includes about 40 images per class and the source of these images are
shopping websites like alibaba.com and indiamart.com etc.
 Tensorflow keras is such an library found in python well suited for CNN’s. the pre-trained model
was available in this library. Apart from that another function called ImageDataGenerator was
available which can generate samples out of images.
 The streamlit was used to develop a web application design format from the developed code.
 The ngrok was used to host a website and create URL for the web application.

WORKING CONCEPT
STEP 1- IMPORTING ESSENTIAL LIBRARIES
This step involves importing libraries required for this project. It includes the VGG16 model,
some libraries for layers in network, libraries for image processing and for web application.

STEP 2- IMPORTING THE DATASET
This step involves importing the dataset. The dataset is basically a folder which consists of
two subfolders namely train and validation. There are the 20 classes in both the train and
validation folders.
The dataset is present in the google drive and hence to import it into the IDE, it is
to import google drive in the IDE and the path for the training and testing data are given.

STEP 3- LOADING THE VGG16
The VGG16 model is stored in the models of keras library. The partial training has been
achieved by keeping the trainable layers as false (layer freezing). About 2 dense (FC) layers
along with dropout and batchnormalization have been added. The model is then compiled
using Adam algorithm and cross entropy loss.

STEP 4- IMAGE PREPROCESSING
This step involves processing the image before introducing to the model. The
ImageDataGenerator of keras produces samples from images by changing properties like
shear, angle and so on. This has been done for the training and testing images.
The preprocess_input preprocesses the images in need for VGG16 (already stored)

STEP 5- SETTING THE DATA
This step is done to alert the machine on the data. The target_size converts the
into the specified dimensions. This must be done so that the model accepts the
Since this is a multiclass classification the class_mode is categorical. These have been
done on both training and testing images.

STEP 6- TRAINING THE MODEL
This step involves the training of the model designed before. Epochs refers to the number
of steps for which the model has to be trained for. The ModelCheckpoint saves the best
model (model which highest accuracy and least model).

STEP 7- MODEL EVALUATION
This step evaluates the model and returns the loss and the accuracy of the model on the
data. A graph has been plotted to display the trend of loss and accuracy over the epochs

https://drive.google.com/file/d/1YubH7T_CXGRyrPZvQcgueKNgM4l70nd4/view?usp=sharing

CONCLUSION
Hence, we have successfully implemented VGGNet to detect and classify the given image
among the 20 common biomedical instruments and in this process we have achieved about
93% accuracy. The model was able to correctly detect the uploaded images of biomedical
instruments of the 20 classes. The model is also able to detect the machine even if the image
of the parts are given (model correctly predicted endoscopy when image of the camera was
uploaded and catheter when the image of the tube was uploaded) .
Accuracy: 93.4%
Time Taken: 5 min and 12 s (depending upon the epochs)

ADVANTAGES
People could use this system (as website or an app) to gain information about several
medical devices surrounding them.
In a short period of time, the required output information is obtained from the
corresponding image input.
Several scans can be classified into various categories depending upon the recording
instrument or technique.
Better accuracy has been obtained (about 93%) on the datasets obtained from online
shopping websites. Can achieve even better with images from hospitals.
Several platforms or sites can use digital image processing as a feature to label various
medical instruments.
This idea could help the healthcare industry by keep a timeline or track of
advancements in several biomedical instruments, devices and techniques.

PROFILE OF THE TEAM
SUBMITTED BY:-
NAME- Arjun Bhattacharya
Dept.- Biomedical Engineering
Year- II Year
e-mail-
arjun.bhattacharya.2019.bme@raj
alakshmi.edu.in
SUBMITTED BY:-
NAME- V. A. Sairam
Dept.- Biomedical Engineering
Year- II Year
e-mail-
sairam.va.2019.bme@rajalakshmi.
edu.in
COLLEGE→ Rajalakshmi Engineering
College

Detection of medical instruments project- PART 2

More Related Content

What's hot (19)

Similar to Detection of medical instruments project- PART 2 (20)

More from Sairam Adithya (9)

Recently uploaded (20)

Detection of medical instruments project- PART 2