🚨 Pulse of Innovation 🚨

Breaking news in the healthcare AI

Trends in investment for HealthTech and AI in 2024 

• The healthcare industry is increasingly adopting administrative automation, workforce management solutions, and generative AI, reflecting a strategic shift towards efficiency and innovation to tackle labor shortages.

• Investment trends within the sector indicate a preference for profitability and cash flow-positive deals, signaling a cautious yet optimistic outlook amidst market corrections.

• There's a notable focus on advancing digital health and value-based care, driven by the need to address workforce challenges, enhance care delivery, and improve patient engagement and outcomes, highlighting the sector's commitment to evolving healthcare paradigms.

Practical AI integration in EHR could save physicians 1 hour daily in documentation.

• News from HIMSS24 shows how eClinicalWorks has implemented several practical AI modules in their EHR system to automate several tasks and save one hour per day in a physician’s life by reducing documentation times.

• Robotic process automation (RPA): EHR vendors are using RPA to create automated "playlists" for quick, multiscreen navigation within the EHR. This lets providers complete repetitive tasks with a single click, saving them valuable time and effort.

• The incoming fax documents are auto-identified by AI-driven image processing and are categorized and saved with corresponding patient charts. Continued integration of such solutions will improve timeworthiness and likely reduce physician burnout.

🧑🏼‍🔬 Bench to Bedside👨🏽‍🔬

Developments in healthcare AI research and innovations

Can Healthcare AI act as a generalist (think of it as a primary care physician of AI)?

• MultiMedBench, a benchmark of 14 diverse biomedical tasks, has been developed to assist in creating generalist biomedical AI systems. Med-PaLM M is a single model that can handle a range of biomedical modalities and tasks, from answering questions to image generation/summarization and calling out genomic variants.

• Med-PaLM M demonstrated the ability to generalize to novel medical concepts, such as tuberculosis detection from chest x-rays, and to perform zero-shot multimodal reasoning. Radiologist evaluation of Med-PaLM M-generated chest x-ray reports showed encouraging results, with clinicians expressing a preference for Med-PaLM M reports in up to 40.50% of cases.

• Editors' Clinical Take: We are not sure a generalist model of AI is prime-time yet! In healthcare, accuracy is essential, and currently, the model is subpar (less than 50% of the time) when clinicians find the output valuable over what radiologists did. Maybe the future is not too far, but we don’t have “one AI tool” that is a real jack of all trades!

AI for sepsis prediction in the Emergency Room

• A deep learning model could predict sepsis in emergency department patients with high accuracy, significantly decreasing mortality rates when implemented in two emergency departments.

• The model also improved other quality metrics, such as adherence to treatment guidelines and reduced organ dysfunction. This suggests that the model's success is due to its ability to identify patients at high risk of sepsis early and to prompt clinicians to intervene. However, more research is needed to confirm these findings.

👩‍⚕️AI in the Clinic 🏥

Real-world and practical use of healthcare AI in clinic

Google NotebookLM as your new virtual research assistant for your articles or books.

• Visit Kingle Notebook and create a new notebook, e.g. “Healthcare AI Articles”. Now, add your articles as source files.

• You can now ask questions, request summaries, and more based solely on your provided documents. The responses will include citations that link directly to uploaded files. You can also have AI-generated prompts to help continue your content exploration and improve your recall and research.

All India Institute of Medical Sciences (AIIMS) in Delhi, India implements AI in clinic for early cancer detection.

• AIIMS, New Delhi, in collaboration with the Centre for Development of Advanced Computing (CDAC), Pune, has launched an AI platform named 'iOncology.ai' aimed at early cancer detection to combat the disease's global impact.

• This platform utilizes advanced AI trained on a dataset of 500,000 radiological and histopathological images of breast and ovarian cancer to analyze medical data with high accuracy. The technology has been implemented at AIIMS and five district hospitals.

🩺 Start-Up Stethoscope 💵 

Trending startups and technologies impacting clinical practice

AI-based wound care diagnostic company Spectral AI received UKCA authorization for Burns care

• Spectral AI focused on medical diagnostics for faster and more accurate treatment decisions in wound care its application is focused on burns care as well as diabetec foot ulcer however, current approval is for burns diagnostic indication with commercialization expected in late 2024.

• Its wound imaging device uses AI-enabled algorithms analyzing real-life patient wounds to distinguish between healthy and damaged tissue within seconds with high level of accuracy, 92%, which is superior to clinical efficacy of 50-75%.

Is the company on its way to becoming an AI healthcare unicorn?

• Arbridge - The AI technology by Abridge streamlines medical note-taking by converting patient-clinician conversations into clinical note drafts in real time.

• Abridge secured a $150 million Series C funding round led by Lightspeed Venture Partners and Redpoint Ventures to enhance its generative AI technology for medical documentation, having raised $212.5 million to date.

🤖Patients First, AI Second🤖

Ethical and regulatory landscape of healthcare AI

To do no harm - and the most good - with AI in healthcare

• The RAISE conference addressed the need for responsible and ethical AI in healthcare. The conference's central goal was to ensure that AI in healthcare primarily benefits patients and addresses current shortcomings in healthcare systems, such as medical errors and access disparities.

• A proposal for four critical aims for AI in healthcare was created: improved patient health, enhanced healthcare experiences, reduced care costs, and support for healthcare workers amidst labor shortages and burnout.