Introduction

Today, an increasing volume of research is being done on the role of AI/ML. However, few players are working on using it to aid Head and Neck Cancer detection using a range of imaging modalities. As per reports by international medical research institutes, the global incidence of this disease continues to rise with more than half a million cases annually. Cancer in the Head is four times more in men than in women. The prognosis remains poor, with a 28 to 60 percent chance of survival at five years, depending upon the stage of detection. Socio Economic problems, high patient population, scarcity of well trained health workers, and inadequate medical infrastructure are major challenges faced in the management of these cancers.
In this context, the latest AI based methods developed by Diatoz Solutions can achieve high degrees of accuracy that can exceed the abilities of human judgement in making data based predictions. So, large scale multi-centric studies are required to aid deployment into real time clinical practice for the benefit of patients and doctors across the world.
DIATOZ aims to provide solutions on how to leverage deep machine learning models for therapeutic response prediction of head and neck cancer and create awareness about the same.

High Risk Factors

- TOBACCO USE - SMOKING - ALCOHOL - EXPOSURE TO HPV - PERSISTENT NECK MASS OR PAIN IN MOUTH THROAT

Use of Deep Learning (DL) in Early Diagnosis

ML is a branch of AI to use computational methods and detect patterns, gather insight and make predictions about new data by using data available. Deep learning (DL) is part of ML in which algorithms are structured to create artificial neural networks with multiple hidden layers.
These methods have also gained significant popularity in recent years due to their achieving relatively high accuracy of prediction.
The solution uses a strong image processing capability powered by machine learning algorithms to effectively analyse the MRI images of different types of head and neck cancer. The model has been trained with intent to predict cancer. The current innovation will be cost and time effective.In addition to the patient this invention will also help doctors to decide the cancer based on the predicted result and enable them to start the right treatment as soon as possible. This process enhances better patient management, increasing chances of survival. AI/ML based prediction increases accuracy and disease outcomes shall assist healthcare professionals in the early detection and planning of patient-specific optimal treatments to reduce the severe impact of the disease.

Advantages of using AI/DL

Technical Feasibility

This Solution is effective as it consists of a micro-service API and a front-end interface to identify top 10 common cancer in head and neck. The result has high accuracy and high recall. It also displays symptoms and suggests possible treatments. It takes MRI images as input and the next release will include feature severity which will help users to know if they need to see a doctor and what's the right treatment, to avoid misguidance too.

Operational Feasibility

The Potential end users of this solution include doctors, patients as well as vendors. As Stakeholders on board to use these plans - Diatoz will provide a social sign in for all.
Technological specifications include the use of a pre pre-trained deep learning model to train the top 10 cancers from head-neck dataset. This is followed by EDA and then fine tuning the data. Thereby adding layers to the deep learning model for better fitting. It has achieved accuracy of 92% and high recall.
As an offshore team (India) of 10 members in the AI/ML vertical with 15+ years of experience this solution is more efficient and will be constantly updated based on new technical developments in the field.

Features to Diagnose Cancer

- Upload MRI image from files to the computer - View the photo to verify - Predicted Cancer, observe and identify the respective symptom and suggest treatment, then easily download the report for sharing with doctors or patients.

Technology Resources used include

- Solution Architect- AI/ML - AI/ML SDE - UI/UX designer - Android developer - QA

Conclusion

This technology solution provides an affordable and efficient solution to screen patients testing them for Head and Neck Cancer.
DIATOZ team is an expert in handling big data, extracting semantics and value propositions using several scientifically proven algorithms. Deep learning tools help predict and propose what could help medical experts make a quick and right decision, to maintain higher rates of success in treating this illness. We also provide end to end solutions in visualising data in various platforms and devices. These services are cost effective since we develop and use sophisticated cloud based services that keep your analytics related costs in control while empowering business with analytics insights.
Read More