-
Large Language Models with Retrieval-Augmented Generation for Zero-Shot Disease Phenotyping
Authors:
Will E. Thompson,
David M. Vidmar,
Jessica K. De Freitas,
John M. Pfeifer,
Brandon K. Fornwalt,
Ruijun Chen,
Gabriel Altay,
Kabir Manghnani,
Andrew C. Nelsen,
Kellie Morland,
Martin C. Stumpe,
Riccardo Miotto
Abstract:
Identifying disease phenotypes from electronic health records (EHRs) is critical for numerous secondary uses. Manually encoding physician knowledge into rules is particularly challenging for rare diseases due to inadequate EHR coding, necessitating review of clinical notes. Large language models (LLMs) offer promise in text understanding but may not efficiently handle real-world clinical documenta…
▽ More
Identifying disease phenotypes from electronic health records (EHRs) is critical for numerous secondary uses. Manually encoding physician knowledge into rules is particularly challenging for rare diseases due to inadequate EHR coding, necessitating review of clinical notes. Large language models (LLMs) offer promise in text understanding but may not efficiently handle real-world clinical documentation. We propose a zero-shot LLM-based method enriched by retrieval-augmented generation and MapReduce, which pre-identifies disease-related text snippets to be used in parallel as queries for the LLM to establish diagnosis. We show that this method as applied to pulmonary hypertension (PH), a rare disease characterized by elevated arterial pressures in the lungs, significantly outperforms physician logic rules ($F_1$ score of 0.62 vs. 0.75). This method has the potential to enhance rare disease cohort identification, expanding the scope of robust clinical research and care gap identification.
△ Less
Submitted 11 December, 2023;
originally announced December 2023.
-
Mapping Physical Conditions in Neighboring Hot Cores: NOEMA Studies of W3(H$_2$O) and W3(OH)
Authors:
Morgan M. Giese,
Will E. Thompson,
Dariusz C. Lis,
Susanna L. Widicus Weaver
Abstract:
The complex chemistry that occurs in star-forming regions can provide insight into the formation of prebiotic molecules at various evolutionary stages of star formation. To study this process, we present millimeter-wave interferometric observations of the neighboring hot cores W3(H$_2$O) and W3(OH) carried out using the NOEMA interferometer. We have analyzed distributions of six molecules that acc…
▽ More
The complex chemistry that occurs in star-forming regions can provide insight into the formation of prebiotic molecules at various evolutionary stages of star formation. To study this process, we present millimeter-wave interferometric observations of the neighboring hot cores W3(H$_2$O) and W3(OH) carried out using the NOEMA interferometer. We have analyzed distributions of six molecules that account for most observed lines across both cores and have constructed physical parameter maps for rotational temperature, column density, and velocity field with corresponding uncertainties. We discuss the derived spatial distributions of these parameters in the context of the physical structure of the source. We propose the use of HCOOCH$_3$ as a new temperature tracer in W3(H$_2$O) and W3(OH) in addition to the more commonly used CH$_3$CN. By analyzing the physically-derived parameters for each molecule across both W3(H$_2$O) and W3(OH), the work presented herein further demonstrates the impact of physical environment on hot cores at different evolutionary stages.
△ Less
Submitted 16 November, 2023; v1 submitted 18 October, 2023;
originally announced October 2023.
-
Comparing Complex Chemistry in Neighboring Hot Cores: NOEMA Studies of W3(H$_{2}$O) and W3(OH)
Authors:
Will E. Thompson,
Morgan M. Giese,
Dariusz C. Lis,
Susanna L. Widicus Weaver
Abstract:
Presented here are NOEMA interferometric observations of the neighboring hot cores W3(H$_{2}$O) and W3(OH). The presence of two star-forming cores at different evolutionary stages within the same parent cloud presents a unique opportunity to study how the physics of the source and its evolutionary stage impact the chemistry. Through spectral analysis and imaging, we identify over twenty molecules…
▽ More
Presented here are NOEMA interferometric observations of the neighboring hot cores W3(H$_{2}$O) and W3(OH). The presence of two star-forming cores at different evolutionary stages within the same parent cloud presents a unique opportunity to study how the physics of the source and its evolutionary stage impact the chemistry. Through spectral analysis and imaging, we identify over twenty molecules in these cores. Most notably, we have detected HDO and CH$_{3}$CH$_{2}$CN in W3(OH), which were previously not detected in this core. We have imaged the molecular emission, revealing new structural features within these sources. W3(OH) shows absorption in a "dusty cocoon" surrounded by molecular emission. These observations also reveal extended emission that is potentially indicative of a low-velocity shock. From the information obtained herein, we have constructed column density and temperature maps for methanol and compared this information to the molecular images. By comparing the spatial distribution of molecules which may be destroyed at later stages of star formation, this work demonstrates the impact of physical environment on chemistry in star-forming regions at different evolutionary stages.
△ Less
Submitted 18 July, 2023;
originally announced July 2023.
-
A Fuzzy Logic Approach to Target Tracking
Authors:
Chin-Wang Tao,
Wiley E. Thompson
Abstract:
This paper discusses a target tracking problem in which no dynamic mathematical model is explicitly assumed. A nonlinear filter based on the fuzzy If-then rules is developed. A comparison with a Kalman filter is made, and empirical results show that the performance of the fuzzy filter is better. Intensive simulations suggest that theoretical justification of the empirical results is possible.
This paper discusses a target tracking problem in which no dynamic mathematical model is explicitly assumed. A nonlinear filter based on the fuzzy If-then rules is developed. A comparison with a Kalman filter is made, and empirical results show that the performance of the fuzzy filter is better. Intensive simulations suggest that theoretical justification of the empirical results is possible.
△ Less
Submitted 13 March, 2013;
originally announced March 2013.
