Category: Uncategorized

Tackling The Cancer Epidemic Using AI

In recent times, breast cancer has become the most common type of cancer affecting women worldwide accounting for 25% of all cancer cases and affected 3.5 million people in 2017-18. Early diagnosis in these cases significantly increases the chances of survival. The key challenge in cancer detection is how to classify tumours into malignant or benign. Research indicates that most experienced physicians can diagnose cancer with 79% accuracy while using artificial intelligence based diagnosis, it is possible to achieve 91% accuracy.

NASSCOM CoE-IoT & AI & Eqounix Tech Lab organised a hands-on session on 28th September 2019, 10am onwards at the CoE Gurugram center on Breast Cancer detection. This was a paid session and there were over 50 attendees, consisting of basic & advanced developers from enterprises like Sopra Steria, United Health Group, TCS, HCL, Inventum Technologies, Publicis Sapient, Globallogic etc and startups like Attentive AI, Empass, Anasakta Labs, NEbulARC, SirionLabs & Vision Networkz. Students from ICGEB/JMI, Indira Gandhi Delhi Technical University etc also participated in the session.

The session first half focussed on Wisconsin Diagnostic Cancer (WDBC) & Invasive Ductal Carcinoma (IDC) datasets insights, Visualization of Dataset, feature selection and why they are chosen and how the physical parameters are translated into a dataset. In the second part, the focus was on Feature Selection and CNN, random forest based classification of cancer as malignant or benign followed by the optimised the deployment strategy & cost estimation

WISCONSIN DIAGNOSTIC BREAST CANCER (WDBC):

The samples consist of visually assessed nuclear features of Fine Needle Aspirates (FNAs) taken from patients. Attributes 3 to 11 were used to form a 9-dimensional vector which was used to obtain a neural network to discriminate between benign and malignant samples. Cross-validation was used to project the accuracy of the diagnostic algorithm.

Field Attribute
1Sample code number
2Class: 2 for benign, 4 for malignant
3Clump Thickness
4Uniformity of Cell Size
5Uniformity of Cell Shape
6Marginal Adhesion
7Single Epithelial Cell Size
8Bare Nuclei
9Bland Chromatin
10Normal Nucleoli
11Mitoses

INVASIVE DUCTAL CARCINOMA (IDC)

Invasive Ductal Carcinoma dataset originally consisted of 162 slide images, scanned at 40x. From that, 277,524 patches of 50×50 pixels (which is converted to  32 x 32pixels to fit the model architecture)  were extracted, including, 198,738 IDC negative examples & 78,786 IDC positive examples. Each image in the dataset is labelled based on the following parameters:

  • Patient ID: 10253_idx5
  • x-coordinate of the crop: 1,351
  • y-coordinate of the crop: 1,101
  • Class label: 0 (0 indicates no IDC while 1 indicates IDC)

CNN Architecture used:

  1. Used exclusively 3×3 CONV filters, similar to VGGNet
  2. Stacked multiple 3×3 CONV filters on top of each other prior to performing max-pooling (again,

similar to VGGNet)

  1. But unlike VGGNet, used depthwise separable convolution rather than standard convolution layers
  2. Keras Sequential API is used to build CancerNet

The model achieved 86% classification accuracy, 85% sensitivity, and 85% specificity.

CODE REPOSITORY:

https://bitbucket.org/tdequonix/cancer_detect_meetup_resources/src/master/

Enter git clone and the repository URL at your command line:

git clone https://shantanugaur@bitbucket.org/tdequonix/cancer_detect_meetup_resources.git

  • build_dataset.py : Builds the dataset by splitting images into training, validation and testing sets.
  • cancernet.py : Contains CancerNet breast cancer classification CNN
  • train_model.py : Responsible for training and evaluation of Keras classification model

DATASET & PRESENTATION: 

https://drive.google.com/open?id=1kgbtXSkDhDXK1aTCb8l1GvQcOn3iTjZQ

NEXT SESSION:

The next session will be focussed on the Prognostic system. This recently put into clinical practice, is a method that predicts when the cancer is likely to recur in patients that have had their cancers excised.  This gives the physician and the patient better information with which to plan treatment and may eliminate the need for a prognostic surgical procedure. The novel feature of the predictive approach is the ability to handle cases for which cancer has not recurred as well as cases for which cancer has recurred at a specific time.

CDAC Workshop on Healthcare Data Standards, Data Privacy and Data Security

Electronic/Personal Health Records (EHR/PHR) and Hospital Information Management Systems (HIMS) are at the core of healthcare data aggregation, as innovators strive to integrate their technology solutions with the country’s existing EHR/PHR and HIMS databases.

C-DAC, an entity of MeitY, is at the helm of innovation in technology for several industries including healthcare. For the last 20+ years C-DAC has been working on Health Information Solutions covering multiple business needs prioritised by the Government.

NASSCOM CoE-IoT organized a workshop where the leadership teams from C-DAC provided an overview of the entire health-stack developed and implemented by them. Nearly 30 Startups participated in the workshop to learn about the relevance of EHR2016 standards, provisions in the draft DISHA guidelines on Data Privacy & Security and the modalities for integrating their digital solutions with C-DAC applications through API integration.

An interactive Panel Discussion was moderated by Raghuram Janapareddy, Director (Lifesciences & Healthcare), NASSCOM CoE with the panellist Mr.Vivek Khaneja Executive Director, Mr.Pavin Srivastava, Associate Director and Mr.Rajiv Yadav Associate Director. The startups clarified their doubts on Data Security/Privacy, EHR Standards and the opportunities to work with C-DAC. Later Startups made presentations about their solutions for the benefit of the C-DAC Team.

NASSCOM CoE-IoT and IIC join hands for promotion of Digital Transformation of Manufacturing Industry

Pune: The combination of advanced manufacturing process and diverse digital technologies continues to proliferate digital transformation of manufacturing enterprises in India, but at the same time it experiences setback and resistance to the adoption. The wireless communications, internet, innovative solutions, and predictive maintenance allows manufacturing industries to benefit from lower adoption costs and increased productivity and efficiency.

The ‘Digital Transformation of the Manufacturing and Heavy Industry’ conference jointly organized by NASSCOM & IIC was held on May 24, 2019 at Sheraton Grand Hotel, in Pune. The objective of the conference was to highlight the impact by creating an ecosystem of technology players, manufacturing enterprises and innovative start-ups.

Dr. Ganesh Natrajan, 5FWorld, Chairman, shared his insight about intelligent factory operations, demand forecasting, intelligent data security solutions and predictive maintenance. He discussed several case studies from Siemens on intelligent gas turbines, GE Aviation – digital twins for predictive maintenance and Jabil predictive factories. Various key factors which could influence the adoption of and define the roadmap of Industry 4.0 in the Indian manufacturing industry were discussed.

Dr. Manoj Athawe, Cummins Group, delivered the keynote session where he discussed the IoT telematics ecosystem and its future in the automotive industry. Dr. Athawe also shared his views on the smart grid battery charging stations and how it could be leveraged to promote renewable energy resources.

Dr. Kiran Bala, SKF, said that the changing employment landscape and key challenges of leadership, workforce skill gap and infrastructure imperative are being faced by every industry. She also stressed on various skills imperative – both digital and cross-functional in nature to be adopted by manufacturing firms in taking a step forward.

During the session three solution providers shared their experience with different solutions that they have deployed and how it benefited the users.

The Altizon team spoke about how they investigate defective manufacturing practices and flawed quality control in their customer operations. The team works on the cause and effect analysis using several 5S quality tools such as Ishikawa diagram to deal with process genealogy & map the complete value chain analysis.

In one of the case studies shared, Intellinet have reduced production time for many customers with the implementation of digital solutions by capturing date from different combination of sensors, PLC and serial interface. Many issues which Intellinet faced during implementation, increased involvement of foot soldiers, addressing maintenance related queries and adopting continual improvement processes to it by appointing local champions on the shop floor and operations and maintenance representative per machine type.

Entrib team identified the challenges faced by one of the automotive two-wheeler OEM. The OEM faced challenges like missing insights on quality improvement, uncontrollable utility costs and limited process monitoring and visibility. The team worked on complete digitalization of the paint shop and various operations with generation of a digital dashboard. The solution generated meaningful insight along with analysis of each process with basic predictive alerts.

The conference also featured a high powered panel discussion, consisting of speakers from both solution providers as well as solution users. Speakers included Mr. Pranab Prasad from Tata Motors, Mr. Anupam Shrivastav from Sudarshan Chemicals, Mr. Shirish Sabnis from John Deere, Mr. Rajeev Dev from Cybage, Mr. Anurag Ambedkar from Maven and Dr. Arvind Tilak from AIPL. The discussion was moderated by Mr. Sudhanshu Mittal from NASSCOM CoE-IoT.

Key items from the panel discussions were around the fact that most of the solution providers working in the area of Analytics / AI / Prediction don’t have domain knowledge when they pitch their solutions to the users and these turn into a long duration co-create projects for users. Also, the challenges in scaling from PoC to actual deployment and even identifying the right solution partners were discussed. Most of the time users don’t have the requisite technical expertise and hence face problem in being able to evaluate the solution providers pitching their solutions. The perspective that solution providers give, was that top management of users are more focused on buzzwords like AI / ML etc., without realizing that each operation is unique and hence not feasible for any solution provider to start doing predictive analysis and suggestions from the first day itself.

In addition, many a times the manufacturing companies have just basic SAP solutions being used for finance / HR work, without any clarity about the ability to handle the incoming data from the shop floor. Most of the times there is no single point of contact to discuss with, when addressing technical challenges. In general, the solution providers do not find it necessary to include the security aspects as the user usually is uninterested and is just looking for quick and inexpensive deployment. There was rigorous exchange with the audience who were interested in further exploration of thoughts from panelists regarding the addressing of the challenges and bridging the gap between the solution providers and the end users.

Pain-free, accurate breast cancer detection? AI, Sir.

India is a goldmine for innovation in healthcare. A billion-strong population, scattered across 29 states, each with its own unique living conditions is bound to create a cornucopia of healthcare challenges. As the country has progressed, so has the landscape of disease. India is a hotbed for non-communicable diseases (NCDs) such as diabetes, cancer and heart ailments. Unlike invasive surgical procedures, treatment of NCDs are time-consuming and expensive.

One of the most common ailments seen today is breast cancer. One in every 28 women are likely to develop breast cancer during their lifetime. India, unfortunately, has among the top incidence rates for this disease and had the highest mortality rate for breast cancer globally in 2017. Ask any oncologist and they will tell you that early detection can go a really long way in mitigating fatality. Moreover, the cost of later detection is almost twice as high compared to early detection.

However, early stage scanning in India conventionally involves mammograms, which exposes a woman to needless amounts of radiation and X-Rays. Additionally, mammograms are generally done on women over the age of 45 and show anomalies in the breast tissue as prominent white spots on an X-Ray.

What about women below the age of 45? And what about benign tumours?

This was a critical inflection point for Geetha Manjunath and Nidhi Mathur, founders of NIRAMAI, one of India’s first AI-powered, non-invasive breast cancer screening devices. NIRAMAI is an acronym for Non Invasive Risk Assessment with Machine Learning & Artificial Intelligence – and uses only a thermal imaging device to measure temperature distribution on the chest. Data is then analysed by a cloud-based analytics platform that assesses if the breast has malignant tumours.

Unlike other screening devices, NIRAMAI is powered by Artificial Intelligence and is a definitive example of digital propulsion to address healthcare challenges in India

Technology at the core of Niramai

At the very heart of NIRAMAI’s solution is an AI-enabled, computer-aided diagnostic engine called Thermalytix, which uses a high resolution thermal sensing device and a cloud-hosted analytics solution to survey images. A combination of big data analytics, AI and ML are used to detect breast cancer at an early stage.

NIRAMAI’s cancer screening tool SMILE has been tested on more than 4,000 women in 12 hospitals, diagnostic centres and screening camps

The results from NIRAMAI’s proprietary platforms indicate:

  • 90% sensitivity in all trials
  • 27% higher accuracy than mammograms
  • 70% higher positive predictive value than visual interpretation of thermography
  • Accurate detection of cancer lesions as small as 4 mm

With the use of technology, NIRAMAI has revolutionized breast cancer detection in India by making it accurate, accessible, non-invasive and completely technology-enabled.

At the core of India’s diverse healthcare spectrum lies a common challenge – uniting the level of care, diagnosis and treatment that can drastically improve the quality of life. Technology is the answer to this challenge, and companies like NIRAMAI are proving this adage to be true.

At the upcoming Lifesciences & Healthcare Innovation Forum (LHIF), scheduled to be held in New Delhi on April 30, 2019, experts from India’s medical and healthcare community, startups, innovators, government and academia will convene to discuss the Digital Propulsion of Universal Healthcare.

To register, visit http://www.coe-iot.com/lhif5/

BFSI – Strategy Innovation Group (BFSI-SIG)

 

 

Mobilizing Industry Insights forApplication of AI in BFSI

 A NASSCOM CoE initiative

 

There was a time in India when going to the bank meant spending at least half a day in a long queue, waiting endlessly for a user’s financial details to be updated in paper passbooks and ledgers. The bank of today is completely different – its digitized, fast and also, smart. Smart is a simpler way of denoting the extremely intuitive and efficient manner in which banks are processing huge amounts of user data. Banks are now providing smart chatbots, personalized services, preventive maintenance, fraud detection and more, to enhance customer experience and improve operational efficiency.

 

Increasingly, the use of AI and ML to various functions within banking and financial services (BFSI) has revolutionised the industry in India.

 

According to data provided by NITI Aayog and Accenture, AI can boost corporate profits by 38% by 2035, potentially adding $1 trillion to the economy. Specifically, AI is expected to have far-reaching impact on the financial services industry on traditional cost bases:

 

  • Banking Industry will witness a $450bn impact by AI
  • Investment Management will witness $200bn in AI impact
  • Insurance Industry will see a $400bn impact thanks to AI

 

In order to affect change within the industry and catalyse a difference in approaching business challenges, new solutions need to be more economically impactful than the current traditional cost base.

Over the past 7-8 years, there has been an immense amount of traction in the Indian fintech industry. Fintechs such as Paytm and Mobikwik brought to India the first wave of innovation by providing mobile wallets and payment options. Slowly, there has been an infusion of fintechs dabbling in provision of financial services such as credit to SMEs, insurance, remittance and accounting. A billion-strong population has enabled the stupendous growth of these fintechs, and VC confidence is testimony to the same. In 2012, VC funding in fintechs stood at $22mn while in 2015, the number rose to $1580mn – the highest to date. At 29%, India offers the highest rate of return on fintech projects compared to the global average of 20%.

 

But now, the time has come to take this burgeoning industry to the next level by leveraging the power of technologies like Artificial Intelligence & Machine Learning. While the innovation in the BFSI industry right now is largely restricted to chatbots, experts believe that this is just the “tip of the iceberg” – and a whole new world of innovation awaits the new-age customer. AI and ML applications can be used to aid decisions on transactions, wealth management, insurance policies, review & monitor financial fraud and so much more.

 

 

In light of accelerating innovation in BFSI, NASSCOM CoE has formed a Strategy Innovation Group (SIG) and first roundtable meeting with thought leaders in the BFSI industry was held on 5th March 2019, where the goal was to mobilize traction in the industry to take innovation to the next level.

 

Highlights of the BFSI-SIG:

 

  1. Mobilize AI and Fintech Startups

Common observations around the group concluded that while innovation is definitely on the upswing with financial startups in India, there needs to be consistency and quality to match up to existing enterprise standards. The reach of India’s BFSI cannot be overlooked, and it would bode well for startups to work with enterprises in a more integrated manner to affect change in a horizontal and vertical manner. The added expertise, wealth of experience and industry connections that enterprises have can massively be leveraged by startups, which are spearheading innovation.

 

  1. Bridge the Gap Between Problems & Solutions

A pertinent observation made by some members of the group was that startups were creating more problem statements to solve, rather than analysing how to provide solutions to existing challenges in the financial sector. The need of the hour is domain-specific, horizontal innovation. In this aspect, NASSCOM CoE can act as a conduit between enterprises and startups to create a repository of existing challenges within the enterprises, and roping in the most appropriate startups to address these challenges.

 

  1. Data Management Critical In FS

Without data, innovation in AI will not happen. However, the rules of data management in India still need to be chalked out properly. With regulatory approvals being put in place in light of growing cyber-crime and financial fraud, having systematic checks and standards in financial data management is critical. With inputs from a cross-section of BFSI industry bodies and decision-makers at the government level, there can be a regulatory data management framework in place for enterprises and startups looking to innovate in this space. NASSCOM CoE can tap into its existing networks and expertise to mobilize data management standards and protocols.

 

  1. Innovation in other tech like Blockchain & IoT

According to some members of the SIG, innovation in areas like blockchain and IoT among startups is yet to take off. These technologies can be hugely influential in navigating the curve of innovation in BFSI. A nudge in this direction could be beneficial for deeptech startups to enhance their working capabilities, and for banks & financial institutions as well who can lend their data and existing infrastructure to support the growth of such technologies. A right step in this direction would be initiate consortium dialogues.

Current Members of BFSI-SIG:

Participant Member Company Name Designation
Deep Thomas Aditya Birla Management Corporation Group Chief Data and Analytics Officer
Manas Dasgupta ANZ Head of Wealth Technology India, ANZ || Chairperson of ICT &3i Committee, BCIC || Innovation Leader & Diversity Champion
Sandeep Varma DBS Bank Head of Innovation, India
Hari RAJAGOPALAN HSBC Senior Vice President – Sales | Global Liquidity & Cash Management
Ramya Padmanabhan HSBC Innovations Lead – Commercial Banking | Global Transformation Centers
Dr. Vadlamani Ravi IDRBT Head, Center of Excellence in Analytics
Venkat Kumar Kotak Mahindra Bank Vice President at Kotak Mahindra Bank, Inventor, Head of Innovation Labs & Fintech Collaborations
Manish Jain KPMG India Partner – Digital Consulting | Fintech | Digital Ledger | Transaction Banking | Innovation
Arun Mehta Royal Bank of Scotland (RBS) Head of Data & Analytics – Digital Engineering Services
Sonal Tivarekar Swiss Re Head Strategic Initiatives & ARC Bangalore | Sr. Vice President | Global Business Solutions
Ramakrishna Vempati Thomson Reuters Lead Innovation
Rajesh Gupta Nasdaq (India) MD
Suresh Iyer Nomura Services India Pvt. Ltd Executive Director, Head of Enterprise Data Management

 

 

 

 

 

Transforming Sports Gear into Smart Devices to Deliver Real-Time Analytics

The Success Story of SeeHow, a NASSCOM CoE Incubatee

Sports analytics, which has been fairly unchartered territory in India, is now rising in popularity. While the global market is pegged at reaching $4.5bn by 2024, the Indian market is playing catch up. Specifically, the data analytics and AI bug seems to have caught cricket aficionados who are now capitalising on advanced technology to understand finer nuances of the sport.

Bangalore-based technology startup SeeHow is transforming traditional sports gear into smart devices using high-end sensors and employ proprietary algorithms to deliver real-time analytics, feedback and coaching advice that enable players to track and improve their game. read more

ICANN Principal Technologist Alain Durand discusses scope of research on Draft Durand Object Exchange

Durand talks about challenges and benefits of IPv6 adoption at NASSCOM CoE

Alain Durand, principal technologist of Internet Corporation for Assigned Names & Numbers (ICANN) paid a visit to the NASSCOM Center of Excellence Gurugram, where there was a vibrant and informative exchange on the potential benefits of establishing standards to the current proprietary systems of firmware download for IOT devices, and the benefits of adopting IPv6 over IPv4.

Alain Durand, principal technologist of ICANN, at NASSCOM CoE Gurugram
The roundtable was attended by representatives from ARICENT, DLINK, CDOT, CDAC and IIT Delhi apart from representation from ERNET, ICANN and CoE. Shri Rahul Gosain, director at the Ministry of Electronics and IT also attended.

read more

India is a thriving hub for AI; Skilled talent key to its growth

How NASSCOM is addressing this issue ground up

The digital progression of every company today is heavily reliant on technology, and Artificial Intelligence (AI) is the current wave that enterprises are riding on. Companies world over have recognized that AI is changing the way business is being done and the very nature of every business enterprise is becoming “DIGITAL FIRST”, making AI crucial not only to their growth but in some cases to their metamorphosis as a digital enterprise, and subsequently, their survival. This is driving them to adopt, adapt and strengthen their AI applicability and fortify their businesses. This inevitably means ensuring AI competence is built and updated continuously.

read more

LHIF Report Nov 2018

Event Report

In an era where everything is available at the click of a button or swipe of a finger, why should the healthcare industry be limited to traditional ways? The need for technology adoption by healthcare industry is more compelling now than ever before. Today, the world of telemedicine and digital healthcare is a coveted field for many startups looking to make their mark in an industry that’s currently on a rise.

(Pronounced as LIFE) – LHIF (Lifesciences & Healthcare Innovation Forum) is a first-of-its-kind initiative by NASSCOM Center of Excellence for AI/IoT to provide a collaborative platform for the healthcare delivery eco-system to explore and adopt innovation in their organizations.

read more

Say hello to #LHIF

The Life Sciences and Healthcare Industry is expected to grow exponentially in India. Federal and State governments across the globe are pushing the pedal on transformational initiatives to achieve these challenging goals. Launch of National Heath Protection Scheme and the intent to establish 150,000 wellness centres under the Ayushman Bharat Yojana, is a major step by the Indian Government towards achieving the UHC goals, presenting major growth opportunities for all the players in Life Sciences & Healthcare Industry.

read more

WordPress Themes