Bailicai: A Domain-Optimized Retrieval-Augmented Generation Framework for Medical Applications

Large Language Models (LLMs) have exhibited remarkable proficiency in natural language understanding, prompting extensive exploration of their potential applications across diverse domains. In the medical domain, open-source LLMs have demonstrated moderate efficacy following domain-specific fine-tuning; however, they remain substantially inferior to proprietary models such as GPT-4 and GPT-3.5. These open-source models encounter limitations in the comprehensiveness of domain-specific knowledge and exhibit a propensity for 'hallucinations' during text generation. To mitigate these issues, researchers have implemented the Retrieval-Augmented Generation (RAG) approach, which augments LLMs with background information from external knowledge bases while preserving the model's internal parameters. However, document noise can adversely affect performance, and the application of RAG in the medical field remains in its nascent stages. This study presents the Bailicai framework: a novel integration of retrieval-augmented generation with large language models optimized for the medical domain. The Bailicai framework augments the performance of LLMs in medicine through the implementation of four sub-modules. Experimental results demonstrate that the Bailicai approach surpasses existing medical domain LLMs across multiple medical benchmarks and exceeds the performance of GPT-3.5. Furthermore, the Bailicai method effectively attenuates the prevalent issue of hallucinations in medical applications of LLMs and ameliorates the noise-related challenges associated with traditional RAG techniques when processing irrelevant or pseudo-relevant documents.

MHLAT: Multi-hop Label-wise Attention Model for Automatic ICD Coding

International Classification of Diseases (ICD) coding is the task of assigning ICD diagnosis codes to clinical notes. This can be challenging given the large quantity of labels (nearly 9,000) and lengthy texts (up to 8,000 tokens). However, unlike the single-pass reading process in previous works, humans tend to read the text and label definitions again to get more confident answers. Moreover, although pretrained language models have been used to address these problems, they suffer from huge memory usage. To address the above problems, we propose a simple but effective model called the Multi-Hop Label-wise ATtention (MHLAT), in which multi-hop label-wise attention is deployed to get more precise and informative representations. Extensive experiments on three benchmark MIMIC datasets indicate that our method achieves significantly better or competitive performance on all seven metrics, with much fewer parameters to optimize.

Jacobian Methods for Dynamic Polarization Control in Optical Applications

Dynamic polarization control (DPC) is beneficial for many optical applications. It uses adjustable waveplates to perform automatic polarization tracking and manipulation. Efficient algorithms are essential to realizing an endless polarization control process at high speed. However, the standard gradientbased algorithm is not well analyzed. Here we model the DPC with a Jacobian-based control theory framework that finds a lot in common with robot kinematics. We then give a detailed analysis of the condition of the Stokes vector gradient as a Jacobian matrix. We identify the multi-stage DPC as a redundant system enabling control algorithms with null-space operations. An efficient, reset-free algorithm can be found. We anticipate more customized DPC algorithms to follow the same framework in various optical systems.

A Pseudo Label-wise Attention Network for Automatic ICD Coding

Automatic International Classification of Diseases (ICD) coding is defined as a kind of text multi-label classification problem, which is difficult because the number of labels is very large and the distribution of labels is unbalanced. The label-wise attention mechanism is widely used in automatic ICD coding because it can assign weights to every word in full Electronic Medical Records (EMR) for different ICD codes. However, the label-wise attention mechanism is computational redundant and costly. In this paper, we propose a pseudo label-wise attention mechanism to tackle the problem. Instead of computing different attention modes for different ICD codes, the pseudo label-wise attention mechanism automatically merges similar ICD codes and computes only one attention mode for the similar ICD codes, which greatly compresses the number of attention modes and improves the predicted accuracy. In addition, we apply a more convenient and effective way to obtain the ICD vectors, and thus our model can predict new ICD codes by calculating the similarities between EMR vectors and ICD vectors. Extensive experiments show the superior performance of our model. On the public MIMIC-III dataset and private Xiangya dataset, our model achieves micro f1 of 0.583 and 0.806, respectively, which outperforms other competing models. Furthermore, we verify the ability of our model in predicting new ICD codes. The case study shows how pseudo label-wise attention works, and demonstrates the effectiveness of pseudo label-wise attention mechanism.

Causes and Corrections for Bimodal Multipath Scanning with Structured Light

Structured light illumination is an active 3-D scanning technique based on projecting/capturing a set of striped patterns and measuring the warping of the patterns as they reflect off a target object's surface. As designed, each pixel in the camera sees exactly one pixel from the projector; however, there are exceptions to this when the scanned surface has a complicated geometry with step edges and other discontinuities in depth or where the target surface has specularities that reflect light away from the camera. These situations are generally referred to multipath where a given camera pixel receives light from multiple positions from the projector. In the case of bimodal multipath, the camera pixel receives light from exactly two positions from the projector which occurs when light bounce back from a reflective surface or along a step edge where the edge slices through a pixel so that the pixel sees both a foreground and background surface. In this paper, we present a general mathematical model and address the bimodal multipath issue in a phase measuring profilometry scanner to measure the constructive and destructive interference between the two light paths, and by taking advantage of this interesting cue, separate the paths and make two separated depth measurements. We also validate our algorithm with both simulation and a number of challenging real cases.