Generative AI Practitioners in Healthcare Prioritize Industry-Specific and Task-Specific Models as Budgets Surge 300%, New John Snow Labs Survey Finds

First Generative AI in Healthcare Survey Uncovers Trends, Challenges, and Best Practices in Generative AI among Healthcare and Life Sciences Practitioners

LEWES, Del., April 23, 2024 (GLOBE NEWSWIRE) — John Snow Labs, the AI for healthcare company, today announced the findings of the inaugural Generative AI in Healthcare Survey. Conducted by Gradient Flow, the research explores the trends, tools, and behaviors around generative artificial intelligence (GenAI) use among healthcare and life sciences practitioners. Findings showed a significant increase in GenAI budgets across the board, with one-fifth of all technical leaders witnessing a more than 300% budget growth, reflecting strong advocacy and investment.

The survey highlights a number of key priorities of practitioners unique to the healthcare industry. A strong preference for healthcare-specific models was a key criterion when evaluating large language models (LLMs). Requiring models to be tuned specifically for healthcare (4.03 mean response) was of higher importance than reproducibility (3.91), legal and reputation risk (3.89), explainability and transparency (3.83), and cost (3.8). Accuracy is the top priority when evaluating LLMs and lack of accuracy is considered the top risk in GenAI projects.

Another key finding is a strong preference for small, task-specific language models. These targeted models are optimized for specific use cases, unlike general-purpose LLMs. Survey results reflected this, with 36% of respondents using healthcare-specific task-specific language models. Open-source LLMs (24%) and open-source task-specific models (21%) follow behind. Proprietary LLMs are less commonly used, whether through a SaaS API (18%) or on-premise (7%).

In terms of how models are tested and improved, the survey highlights one practice that addresses both the accuracy and compliance concerns of the healthcare industry: human-in-the-loop workflows. This was by far the most common step taken to test and improve LLMs (55%), followed by supervised fine-tuning (32%), and interpretability tools and techniques (25%). A human-in-the-loop approach enables data scientists and domain experts to easily collaborate on training, testing, and fine-tuning models to their exact needs, improving them over time with feedback.

“Healthcare practitioners are already investing heavily in GenAI, but while budgets may not be a top concern, it’s clear that accuracy, privacy, and healthcare domain expertise are all critical,” said David Talby, CTO, John Snow Labs. “The survey results shine the light on the importance of healthcare-specific, task-specific language models, along with human-in-the-loop workflows as important techniques to enable the accurate, compliant, and responsible use of the technology.”

Finally, the survey explores the large amount of remaining work in applying responsible AI principles in healthcare GenAI projects. Lack of accuracy (3.78) and legal and reputational risk (3.62) were reported as the most concerning roadblocks. Worse, a majority of GenAI projects have not yet been tested for any LLM requirements cited. For those that have, fairness (32%), explainability (27%), private data leakage (27%), hallucinations (26%), and bias (26%) ranked as the most commonly tested. This suggests that no aspect of responsible AI is being tested by more than a third of organizations.

An upcoming webinar taking place at 2pm ET on April 30 with Drs. Ben Lorica of Gradient Flow and David Talby of John Snow Labs, provides additional details and analysis of the survey results and the current state of GenAI in healthcare.

About John Snow Labs
John Snow Labs, the AI for healthcare company, provides state-of-the-art software, models, and data to help healthcare and life science organizations put AI to good use. Developer of Spark NLP, Healthcare NLP, the Healthcare GPT LLM, the Generative AI Lab No-Code Platform, and the Medical Chatbot, John Snow Labs’ award-winning medical AI software powers the world’s leading pharmaceuticals, academic medical centers, and health technology companies. Creator and host of The NLP Summit, the company is committed to further educating and advancing the global AI community.

 

Your Guide to Surveys for The Construction Industry

Architects, local authorities and structural engineers use surveys and site scan to provide in-depth information about the condition of a site, its unique features and topographic information. This is so that any proposed construction project can move forward with a clear understanding of an area. Continue reading “Your Guide to Surveys for The Construction Industry”

Japanese City Using GIS on iPads for Field Survey

The City of Tamba in Hyogo prefecture, Japan, performs field survey work using a mobile GIS application designed for Apple iPad tablet computers. The application, called Mobile Matilda, is developed by Osaka based GIS solution specialist, Tsukasa Consulting, using the TatukGIS Developer Kernel 11 (for Delphi edition) with the Embarcadero FireMonkey framework. The mobile application compliments desktop and web-based versions of the Matilda software that are enabled with more features.

Unlike web applications, Mobile Matilda can operate in a local (off-line) environment because map layers and data are read from the iPad memory. Users can update records in the field for later synchronization to the desktop or server via a secure Wi-Fi or GSM connection, when available. TatukGIS development tools enable efficient storage of even huge GIS datasets as SQLite database layers on iOS and Android devices.

A mobile map project can also incorporate information streamed via the web, such as TatukGIS hosted OpenStreetMap web tiles as a background layer.

Field surveys conducted by the Tamba government often involve assets such as vacant houses and solar power generation facilities. In recent years, unoccupied houses left unmanaged have become a major issue for the prevention of crime and disasters, requiring municipal governments to better understand and manage this problem. Investments in solar power facilities must be verified and inventoried to qualify for government financial subsidies.

In the past, field surveys required personnel to carry a camera, a map, and paper-based inventory records into the field to collect information, and then return to the office to associate newly collected data with already existing information. The process was labor intensive and prone to errors. The resulting data was not always usable in an efficient manner. With the Mobile Matilda application, observed through ipad survey software information, photographs, and map figures are electronically associated to the map while on site. For improved workflow, survey assignments can be set up in advance to enable the field researcher to access relevant records simply by tapping and selecting on the tablet computer. The iPad’s GPS information can be used to easily center the map (or aerial photography layer) to the user’s present position.

A user can enter new vector points which can be edited such as to change its position, add attribute information, or specify a direction (e.g., viewing angle). Photos can be added using the iPad’s built-in camera or a linked 360-degree panoramic camera (such as seen at: https://theta360.com/en/).

Learn more about the Tsukasa Consulting and its Mobile Matilda application at: https://www.tsukasa-consulting.net

Learn more about TatukGIS tools for desktop, mobile, and web GIS application development at: www.tatukgis.com

Virtual Surveyor Offers Enhanced Functionality in Free Version of Drone Surveying Software

Virtual Surveyor Offers Enhanced Functionality in Free Version of Drone Surveying Software

AARSCHOT, Belgium, 8 September 2020 – In release 7.4, Virtual Surveyor has added new functionality to the free pricing level of its popular drone surveying software. The “Valley” plan of Virtual Surveyor now allows users to create engineering CAD models from drone orthophotos and/or LiDAR data sets.

“Originally offered as a free viewing tool for imagery and elevation models captured by unmanned aerial vehicles (UAVs), our Valley software plan now enables surveyors to create topographic line surveys for small-area projects,” said Tom Op ‘t Eyndt, Virtual Surveyor CEO.

Available under three pricing plans, the Virtual Surveyor software enables surveyors to derive topographic information from drone data that engineers need for construction, mining, and other excavation projects. The software has become popular among surveyors and engineers because its tools allow them to work effectively with drone data.

Virtual Surveyor generates an interactive onscreen environment where the surveyor selects survey points and breaklines to create a highly accurate lightweight CAD model for further use in engineering design packages. When the surveyor wants to calculate volumetric change directly in Virtual Surveyor using all data points, like stockpiles or a cut/fill map, the software creates a clean terrain model to remove conveyors and machines through simple clicking.

In addition to upgrading the free Valley plan, Virtual Surveyor 7.4 has changed the names and enhanced capabilities of its other purchase plans. The three plans are:

Valley (Free) – Create CAD models from drone images or LiDAR for topographic surveying applications in small areas.

Ridge – Create advanced topographic and quantity surveys, including digital surfaces and contours, as well as stockpile volume calculations, from drone data for advanced surveying projects.

Peak – Designed for large-area projects, this plan includes Productivity Tools, such as intelligent gridding and guided breakline generation that speed processing. Most importantly, Peak enables surveyors to measure and track cut-and-fill progress at construction, mining, and other excavation sites over time.

“All Virtual Surveyor plans now empower surveyors to extract meaningful topographic information from their drone data,” said Op ‘t Eyndt.

Virtual Surveyor 7.4 also includes other functionality improvements that span all three pricing levels, making the software faster and easier to use. Most notably, users can now easily rename layers and items, and move items between layers through drag-and-drop. Download the free Valley application today and see what else is new in the Virtual Surveyor smart drone mapping package.

Current subscribers to Virtual Surveyor will see their software being updated to Version 7.4 automatically. To start a free 14-day trial of Virtual Surveyor and to view details of the three pricing plans, visit www.virtual-surveyor.com.

About Virtual Surveyor

Virtual Surveyor software enables professional surveyors to generate accurate topographic end products from drone imagery five times faster than traditional field work. Now used in 68 countries, Virtual Surveyor allows surveyors to topographic information from drone data in very short timeframes. In a typical application, the suite takes the orthophotos and digital surface models (DSM) extracted from UAV imagery with photogrammetric software like Pix4D, Metashape or DJI Terra and generates an interactive virtual environment onscreen where surveyors can select the survey points and breaklines that define topography. Standard topographic outputs from the Virtual Surveyor software are Surfaces or TINs (Triangular Irregular Network), Contours, Line Surveys and Cut-and-Fill Maps. As such, Virtual Surveyor bridges the gap between UAV photogrammetric processing software and engineering design packages.

Okeanus Science & Technology Introduces ASVs to the Global Rental Market

Okeanus Science & Technology (Okeanus), an established provider of oceanographic rental equipment and turnkey survey systems, today announced a new partnership with SeaRobotics Corporation, a leading developer of unmanned marine systems, in a move to make Autonomous Surface Vehicles (ASVs) increasingly available for international rental.

Continue reading “Okeanus Science & Technology Introduces ASVs to the Global Rental Market”

Poll – Share Your Comments on OpenGov and OpenData

We’d like to hear from you with a little feedback and comments regarding Open Government and OpenData. We’d like you to share your favorite resources, comment on the state of OpenGov and provide some general feedback… enjoy, and thanks for participating! Continue reading “Poll – Share Your Comments on OpenGov and OpenData”

Exit mobile version