著者

本田 卓士 松永 力 金谷 健一

雑誌

研究報告コンピュータビジョンとイメージメディア(CVIM)

巻号頁・発行日

vol.2012, no.18, pp.1-8, 2012-11-26

---

空間をわずかに移動する複数の点の移動前後の位置を 3 次元センサーで計測し,どのような並進,回転,スケール変化が生じているのか,あるいは生じていないのかを判断するモデル選択のために,誤差のある 3 次元データにさまざまな運動モデルを最適に当てはめる新しい方法を提案する.これは, 3 次元アフィン変換の部分群が変数にさまざまな内部拘束を指定して得られることに着目して,内部拘束をもつ 3 次元アフィン変換を拡張 FNS 法によって計算するものである.これにより,従来のように運動ごとに別々のパラメータを導入する必要がなく,すべての部分群が同一の方法で計算できる.この手法をステレオ視による 3 次元シミュレーションデータに対する幾何学的 AIC,幾何学的 BIC,幾何学的 MDL を用いたモデル選択に応用する.Given 3-D sensing data of points slightly moving in space, we consider the problem of discerning whether or not translation, rotation, and scale change take place and to what extent. For this purpose, we propose a new method for fitting various motion models to 3-D noisy data. Based on the observation that subgroups of the 3-D affine transformations are defined by imposing various internal constraints on the variables, our method fits 3-D affine transformations with internal constraints using the scheme of EFNS, which, unlike conventional methods, dispenses with particular parameterizations for particular motion models. We apply our method to simulated stereo vision data and show how model selection using the geometric AIC, the geometric BIC and the geometric MDL works.

著者

本田 卓 針谷 正祥

出版者

東京女子医科大学学会

雑誌

東京女子医科大学雑誌 (ISSN:00409022)

巻号頁・発行日

vol.93, no.4, pp.93-98, 2023-08-25 (Released:2023-08-25)

参考文献数

8

---

Genome wide association study (GWAS) is a type of genetic analysis that looks for association between a particular disease or trait and single nucleotide polymorphism (SNP). The polygenic risk score is the sum of the number of risk alleles weighted the GWAS-estimated effect sizes of each SNP on a disease. Researchers have been exploring the use of polygenic risk score (PRS) to predict disease risk and personalize treatment plans, an approach known as precision medicine. Our study was the first to demonstrate that a PRS based on GWAS data for rheumatoid arthritis (RA) onset can also predict joint damage progression. In particular, we found that the PRS is more accurate for predicting joint damage progression in young-onset patients with RA. To facilitate large-scale validation of our findings, we developed an artificial intelligence-based joint damage scoring system. This system will enable us to further investigate the relationship between PRS and disease severity in a larger, more diverse population. Further research is needed to refine the PRS construction method, particularly in terms of identifying the most informative SNPs and optimizing the weighting scheme for risk alleles.