EHR Integrated Clinical Decision Support Support (iCDSS) for Skin Cancer Detection

December 6, 2023 | 4 minute read
Text Size 100%:

Skin cancer can spread and develop into other types of cancers, affecting organs such as the brain, liver, and kidneys. It’s the 17th most common cancer globally. If it's not identified early, the survival rate can drop below 50%. However, when melanoma is detected promptly, its five-year survival rate is an impressive 99%.

The worldwide market for skin cancer therapies was valued at $7.2 billion in 2021. Forecasts show that this figure might escalate to $14.5 billion by 2031, with a compound annual growth rate (CAGR) of 7.3%, spanning from 2022 to 2031. This figure underscores the pressing need for heightened awareness and improved skin cancer screening. The global populace is in dire need of an affordable, convenient solution that facilitates swift, economical, and straightforward skin cancer diagnosis, enabling timely surgical intervention before the disease advances.

However, the path is fraught with challenges, including public unawareness of skin cancer symptoms, a general hesitation towards routine medical examinations, limited access to diagnostic centers, a shortage of trained professionals combined with a strained healthcare infrastructure, particularly in less developed nations, and underreporting of cases in regions with weaker healthcare setups. Another significant hurdle is the isolated nature of patient records, which hampers a holistic understanding for both medical practitioners and researchers.

In this article, we explore the robust capabilities of Oracle Cloud Infrastructure (OCI) Artificial Intelligence (AI) Services, emphasizing its potent use in thoroughly analyzing patient data for precise skin cancer detection, all at a simple click. We introduce a prototype clinical decision support system (CDSS) that's seamlessly integrated with an electronic health record (EHR). This system provides users with effortless access to the application, making the skin cancer diagnostic process quick, cost-effective, and user-centric. With early detection facilitated by this setup, patients can receive timely surgical treatments, curtailing the progression of the ailment. Additionally, the blog features a demo link, illustrating the straightforward development process of such a CDSS using OCI AI Services.

Integrated solutions with OCI

An integrated CDSS combined with an EHR can use Oracle OCI AI Services to thoroughly examine patient data, providing highly precise skin cancer detection at your fingertips. The implications for both patients and organizations include the following examples:

  • Allowing individuals to identify potential risks
  • Giving healthcare providers tools to assess and address risks
  • Facilitating collaborative care through data sharing across various healthcare entities through CDSSes
  • Enhancing and bolstering research initiatives
  • Simplifying intricate decision-making procedures for medical institutions

The usual user experience with an integrated CDSS for skin cancer detection consists of the following processes:

  • Finding the application through diverse platforms, such as social media, search engines, or app marketplaces
  • Delving into its features, understanding the underlying technology, and examining project specifics
  • Trying out the system by uploading a picture of a mole or skin tag, and promptly receiving results
  • Interacting with the chatbot, raising queries, expressing worries, and considering feedback from medical professionals

The design of this integrated CDSS illustrates how various components interrelate, centering around OCI Vision. An end user goes through the following process:

  • An image is uploaded through a web application.
  • This image gets stored in OCI Object Storage.
  • Detecting the new file, OCI Events activates a Python function, which makes a REST API call to OCI Vision.
  • The AI Vision service accesses the image from Object Storage and conducts its analysis.
  • Analysis results are then displayed on the web application.
  • The user uses Oracle’s Digital Assistant to engage with customer service for scheduling appointments, addressing questions, or enhancing their overall experience.

The machine learning (ML) model behind this process is designed to categorize skin lesions into eight distinct classes, including melanoma, melanocytic nevus, basal cell carcinoma, actinic keratosis, benign keratosis, dermatofibroma, vascular lesion, and squamous cell carcinoma.

A diagram of the integrated CDSS

The solution is built using the following Oracle AI services. Oracle's AI services provide pretrained models that you can custom-train with data to improve model quality, making it easier for developers to adopt and use AI technology.

  • OCI Vision, an AI perpetual service, offers deep-learning–based image analysis at scale.
  • OCI Data Science makes it easy to train and manage machine learning models.
  • OCI Data Labeling provides labeled datasets to more accurately train AI and machine learning models.
  • Oracle Digital Assistant, a platform for the development of intelligent digital assistants and chatbots, integrates with your back-end services
  • Oracle Visual Builder, for progressive application development.

 The return on investment (ROI) is notably high, characterized by the following features:

  • Increased efficiency, cost savings, broader accessibility, and heightened awareness
  • Comprehensive visibility for various user profiles
  • Enhanced experiences for patients, providers, and hospitals
  • A consolidated view, fostering informed decision-making


In this post, we delved deep into the powerful functionalities of OCI AI services, highlighting its significance in the meticulous analysis of patient data to accurately diagnose skin cancer with just a single click. We discussed the user experience, the app's design and framework, and the integration of OCI AI services in its construction.

Now, you can see for yourself what a simplified development journey of this kind of CDSS using OCI AI Services looks like with  this demo.

For access to the source code and any further inquiries, contact Vivek Acharya.

Try Oracle Cloud Free Trial! A 30-day trial with US$300 in free credits gives you access to Oracle Cloud Infrastructure Data Science service. 

To learn more about Oracle AI, visit the following pages:



Vivek Acharya

A passionate digital transformation architect, author/writer, AI & Healthcare evangelist  and a blogger. Responsible for promoting digital transformation; modernizing core business systems, and deriving value. Fifteen plus years of experience in problem solving in the realm of architecture, design, mentoring, C- suite relationships, and navigating projects through stakeholders.

Vivek is pursuing an MBA from Boston University, holds a postgraduate certification on Artificial Intelligence and Machine Learning from UT Austin, digital transformation certificate from Cornell and is a certified IBM Design Thinking coach. Also certified from MIT and Stanford on "Healthcare and AI".

Previous Post

Quantize and deploy Llama 2 70B on cost-effective NVIDIA A10 Tensor Core GPUs in OCI Data Science

Tzvi Keisar | 6 min read

Next Post

Deploy LangChain applications as OCI model deployments

Lu Peng | 5 min read