Résumé
Abstract The objective of this research is to present a comparative analysis using various lengths of time windows (TW) during emotion recognition, employing machine learning techniques and the portable wireless sensing device EPOC+. In this study, entropy will be utilized as a feature to evaluate the performance of different classifier models across various TW lengths, based on a dataset of EEG signals extracted from individuals during emotional stimulation. Two types of analyses were conducted: between-subjects and within-subjects. Performance measures such as accuracy, area under the curve, and Cohen's Kappa coefficient were compared among five supervised classifier models: K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Logistic Regression (LR), Random Forest (RF), and Decision Trees (DT). The results indicate that, in both analyses, all five models exhibit higher performance in TW ranging from 2 to 15 seconds, with the 10 seconds TW particularly standing out for between-subjects analysis and the 5-second TW for within-subjects; furthermore, TW exceeding 20 seconds are not recommended. These findings provide valuable guidance for selecting TW in EEG signal analysis when studying emotions.
Resumen El objetivo de esta investigación es presentar un análisis comparativo empleando diversas longitudes de ventanas de tiempo (VT) durante el reconocimiento de emociones, utilizando técnicas de aprendizaje automático y el dispositivo de sensado inalámbrico portátil EPOC+. En este estudio, se utilizará la entropía como característica para evaluar el rendimiento de diferentes modelos clasificadores en diferentes longitudes de VT, basándose en un conjunto de datos de señales EEG extraídas de individuos durante la estimulación de emociones. Se llevaron a cabo dos tipos de análisis: entre sujetos e intra-sujetos. Se compararon las medidas de rendimiento, tales como la exactitud, el área bajo la curva y el coeficiente de Cohen's Kappa, de cinco modelos clasificadores supervisados: K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Logistic Regression (LR), Random Forest (RF) y Decision Trees (DT). Los resultados indican que, en ambos análisis, los cinco modelos presentan un mayor rendimiento en VT de 2 a 15 segundos, destacándose especialmente la VT de 10 segundos para el análisis entre los sujetos y 5 segundos intrasujetos; además, no se recomienda utilizar VT superiores a 20 segundos. Estos hallazgos ofrecen una orientación valiosa para la elección de las VT en el análisis de señales EEG al estudiar las emociones.
Résumé
Objective To explore the feasibility and operability in identifying the therapeutic dominant stages of traditional Chinese medicine(TCM)based on subdivision model of disease course.Methods The hierarchical Bayesian model was used to differentiate the disease course of 125 cases of premature ovarian failure(POF),and the disease course of POF were divided into the occult stage,diminished ovarian reserve(DOR)stage,premature ovarian insufficiency(POI)stage,and POF stage.An then the paired sample t-test,Pearson correlation analysis and expert in-depth interview were used for the analysis of the therapeutic effects of TCM for POF at various stages.Results(1)Compared with POF stage,DOR and POI stages were frequently intervened by Chinese patent medicine.(2)In DOR(complicated with POI)stage and POF stage,there was significant difference between the degree of TCM intervention and the therapeutic effect(t =-3.70,P＜0.001).(3)The degree of TCM intervention was positively correlated with treatment outcomes in the DOR stage(r = 0.679,P＜0.001),so did in the POF stage(r = 0.432,P＜0.001),but the correlation in the POF stage was slightly lower than that in the DOR stage.(4)The results of in-depth interviews with experts of TCM gynecology showed that in the concealed phase of POF,the prognosis would be most favorable if TCM regulation and intervention were performed.In the DOR stage and POI stage,treatment with Chinese medicine prescriptions usually brought about better curative effect and prognosis.For the patients at POF stage,the therapeutic effect of TCM depended on the patients'compliance and the treatment course,and the effect was relatively not as good as that of the previous stages.Conclusion In the DOR stage and POF stage,the higher the degree of TCM intervention,the better the prognosis will be achieved for the patients treated with western medicine.In the POF stage,the efficacy of TCM intervention is reduced to a certain extent compared with the DOR stage.The results indicated that it is feasible and operable to identify the TCM therapeutic dominant stages based on the subdivision model of disease course.
Résumé
The recording and analysis of activities of calcium signals in neurons is of critical importance in the field of neuroscience.Over the past three decades,various fluorescent calcium imaging techniques not only have been used in the imaging study of functional activities of neuronal communities,but also can be combined with specific markers to record the functional activities of specific types of neuronal communities.To analyze neural activities at the cellular level,a series of preprocessing such as motion correction,cell body recognition,calcium signal extraction and peak deconvolution is required for the collected video.However,current methods for manual preprocessing are time-consuming and laborious,so computer automatic analysis technology is urgently needed to quickly repair the jitter in the video,identify the position and outline of a single cell,extract its activity trajectory and infer the action potential peak.In this paper,the methods of calcium imaging data processing used in recent years are summarized,and the future developments are predicted.
Résumé
Objective The objective of this study is to improve the accuracy of automatic identification in complex background herbal slice images.The goal is to achieve accurate recognition of herbal slice images in the presence of complex backgrounds.Methods The experiment was conducted on a collected and organized dataset of Tibetan herbal slice images.The RGB,HOG,and LBP features of the slices were analyzed.An improved HOG algorithm was used to fuse multiple features,and a deep learning network was utilized for image recognition.Results The proposed method of multi-feature fusion combined with deep learning achieved an identification accuracy of 91.68%on 3610 Tibetan herbal slice images with complex backgrounds.Furthermore,the average identification accuracy for 20 common traditional Chinese medicine slices,such as Chuan Beimu,Hawthorn,and Pinellia,reached 98.00%.This method outperformed existing methods for identifying herbal slices in complex backgrounds,indicating its feasibility and wide applicability for the identification of other traditional Chinese herbal medicines.Conclusion The fusion of multiple features effectively captures distinguishing characteristics of herbal slices in complex backgrounds.It exhibits high recognition rates for Tibetan herbal slices with complex and heavily occluded backgrounds,and can be successfully applied to the recognition of natural scene-based traditional Chinese herbal medicines and herbal slices.This approach shows promising prospects for practical applications.
Résumé
Objective To Explore the feasibility and value of deep learning technology for quality control of echocardiography images.Methods A total of 180985 echocardiography images collected from Sichuan Provin-cial People's Hospital between 2015 and 2022 were selected to establish the experimental dataset.Two task models of the echocardiography standard views quality assessment method were trained,including intelligent recognition of seven types of views(six standard views and other views)and quality scoring of six standard views.The predictions of the models on the test set were compared with the results of the sonographer's annotation to assess the accuracy,feasibility,and timeliness of the runs of the two models.Results The overall classification accuracy of the stan-dard views recognition model was 98.90%,the precision was 98.17%,the recall was 98.18%and the F1 value was 98.17%,with the classification results close to the expert recognition level;the average PLCC of the six standard views quality scoring models was 0.933,the average SROCC was 0.929,the average RMSE was 7.95 and the average MAE was 4.83,and the prediction results were in strong agreement with the expert scores.The single-frame inference time after deployment on the 3090 GPU was less than 20 ms,meeting real-time requirements.Conclusion The echocardiography standard views quality assessment method can provide objective and accurate quality assessment results,promoting the development of echocardiography image quality control management towards real-time,objective,and intelligent.
Résumé
Objective To study the effects of the intelligent hearing-assistive system incorporated in Nuro-tron cochlear implants(CI),including the autonomic acoustic scene recognition(ASR),intelligent strategy config-uration as well as the objective and subjective hearing improvements on recipients.Methods ① To evaluate the per-formance of the ASR matule,in a sound-proof room,the preset five kinds of test audios,including speech,noise,speech in noise,pure music(without human voice)and non-pure Music(with human voice)were played.Each type of scenes included 6 to 9 5 min test files.The prediction accuracy and scene switching times were calculated.② In order to evaluate the noise-reduction performance of the ABeam technology in the speech enhancement module,13 Nurotron? CI recipients were recruited and their speech recognition rate when ABeam was"ON"and"OFF"with noise coming from 90°,180°or 270°were tested,individually.Also,their subjective hearing feedback was evaluated through visual analogue scale(VAS)evaluation.Results The ASR module achieved high prediction performance,with prediction accuracy 99％±4％,96％±9％,94％±12％,94％±15％,92％±13％for speech,noise,noisy speech,pure music and non-pure music,respectively.The scene transation times for each individual scene were 1.1 ±0.3,1.4±0.7,1.3±0.5,1.4±0.8 and 1.3±0.5,indicating that the prediction was also stable.When noise came from the sides and behind of recipients and speech signal from the front,the adaptive dual microphone noise re-duction algorithm ABeam significantly increased the speech recognition score(SRS)in 5 dB signal-to-noise(SNR)environment(P＜0.001),with an average increase of 15.92％.Especially when the noise came from 180 degree backward,the SRS increased 28.68％when ABeam was"0N",which was significantly higher than when ABeam was"OFF"(P＜0.01).Conclusion The intelligent hearing-assistive system can help CI recipients automatically configure appropriate SPSs under different environments,improving the speech intelligibility and hearing comfort.
Résumé
BACKGROUND:MRI is important for the diagnosis of early knee osteoarthritis.MRI image recognition and intelligent segmentation of knee osteoarthritis using deep learning method is a hot topic in image diagnosis of artificial intelligence. OBJECTIVE:Through deep learning of MRI images of knee osteoarthritis,the segmentation of femur,tibia,patella,cartilage,meniscus,ligaments,muscles and effusion of knee can be automatically divided,and then volume of knee fluid and muscle content were measured. METHODS:100 normal knee joints and 100 knee osteoarthritis patients were selected and randomly divided into training dataset(n=160),validation dataset(n=20),and test dataset(n=20)according to the ratio of 8:1:1.The Coarse-to-Fine sequential training method was used to train the 3D-UNET network deep learning model.A Coarse MRI segmentation model of the knee sagittal plane was trained first,and the rough segmentation results were used as a mask,and then the fine segmentation model was trained.The T1WI and T2WI images of the sagittal surface of the knee joint and the marking files of each structure were input,and DeepLab v3 was used to segment bone,cartilage,ligament,meniscus,muscle,and effusion of knee,and 3D reconstruction was finally displayed and automatic measurement results(muscle content and volume of knee fluid)were displayed to complete the deep learning application program.The MRI data of 26 normal subjects and 38 patients with knee osteoarthritis were screened for validation. RESULTS AND CONCLUSION:(1)The 26 normal subjects were selected,including 13 females and 13 males,with a mean age of(34.88±11.75)years old.The mean muscle content of the knee joint was(1 051 322.94±2 007 249.00)mL,the mean median was 631 165.21 mL,and the mean volume of effusion was(291.85±559.59)mL.The mean median was 0 mL.(2)There were 38 patients with knee osteoarthritis,including 30 females and 8 males.The mean age was(68.53±9.87)years old.The mean muscle content was(782 409.18±331 392.56)mL,the mean median was 689 105.66 mL,and the mean volume of effusion was(1 625.23±5 014.03)mL.The mean median was 178.72 mL.(3)There was no significant difference in muscle content between normal people and knee osteoarthritis patients.The volume of effusion in patients with knee osteoarthritis was higher than that in normal subjects,and the difference was significant(P<0.05).(4)It is indicated that the intelligent segmentation of MRI images by deep learning can discard the defects of manual segmentation in the past.The more accuracy evaluation of knee osteoarthritis was necessary,and the image segmentation was processed more precisely in the future to improve the accuracy of the results.
Résumé
Video-based intelligent action recognition remains challenging in the field of computer vision.The review analyzes the state-of-the-art methods of video-based intelligent action recognition,including machine learning methods with handcrafted features,deep learning methods with automatically extracted features,and multi-information fusion methods.In addition,the important medical applications and limitations of these technologies in the past decade are introduced,and the interdisciplinary views on the future application to improve human health are also shared.
Résumé
Objective To present a named entity recognition method referred to as BioBERT-Att-BiLSTM-CRF for eligibility criteria based on the BioBERT pretrained model.The method can automatically extract relevant information from clinical trials and provide assistance in efficiently formulating eligibility criteria.Methods Based on the UMLS medical semantic network and expert-defined rules,the study established medical entity annotation rules and constructed a named entity recognition corpus to clarify the entity recognition task.BioBERT-Att-BiLSTM-CRF converted the text into BioBERT vectors and inputted them into a bidirectional long short-term memory network to capture contextual semantic features.Meanwhile,attention mechanisms were applied to extract key features,and a conditional random field was used for decoding and outputting the optimal label sequence.Results BioBERT-Att-BiLSTM-CRF outperformed other baseline models on the eligibility criteria named entity recognition dataset.Conclusion BioBERT-Att-BiLSTM-CRF can efficiently extract eligibility criteria-related information from clinical trials,thereby enhancing the scientific validity of clinical trial registration data and providing assistance in the formulation of eligibility criteria for clinical trials.
Résumé
Objective To establish the HPLC fingerprint of Rosae Rugosae Flos;To provide references for the quality evaluation of Rosae Rugosae Flos.Methods The HPLC analysis was carried on a COSMOSIL 5C18-MS-Ⅱ column(4.6 mm×250 mm,5 μm);the mobile phase was 2.5 % acetonitrile + 0.1 % formic acid aqueous solution(A)-acetonitrile + 0.1 % formic acid(B)with gradient elution at the flow rate of 0.5 Ml/min;the column temperature was 40℃;the detection wavelength was 350 nm.The similarity of 13 batches of samples was evaluated by Similarity Evaluation System for Chromatographic Fingerprint of TCM(2012 edition).Qualitative analysis was carried out by LC-MS technology.The overall quality evaluation of Rosae Rugosae Flos was carried out by combining clustering analysis,principal component analysis and orthogonal partial least square discrimination.Results The common mode of HPLC fingerprints of Rosae Rugosae Flos was established,and the similarity of 13 samples was good.9 compounds were identified preliminary.13 batches of samples were aggregated into 3 categories by chemical pattern recognition.Conclusion The fingerprints of Rosae Rugosae Flos established in this study combines with chemical pattern recognition method,which has high sensitivity and strong specificity,can provide a basis for the quality evaluation of Rosae Rugosae Flos.
Résumé
Objective:To investigate the recognition of the post competency index system among rural general practice assistant physicians and its influencing factors.Methods:This study was a cross-sectional survey. A questionnaire survey on the recognition of post competency index system was conducted from October 2020 to September 2021 among rural general practice assistant physicians from 10 provinces/municipalities selected by stratified cluster sampling method. The recognition of rural general practice assistant physicians at all levels of indexs and the factors influencing recognition were analyzed.Results:A total of 1 123 questionnaires were distributed and 1 024 valid ones were collected with a recovery rate of 91.18%. Of the 1 024 respondents, 529 were male(51.7%) and 435 were aged 40-49 years(42.5%), which was the highest proportion by age group. The average overall recognition score of the index system was 4.41, and the scores of the primary indexes were 4.32-4.45. Three primary indexes had the highest recognition scores: professional competence, basic health care services, and interpersonal communication and teamwork. The recognition scores on the second level index were 4.18-4.61, and the proportion of recognition scores greater than 4 was over 80%. There were significant differences in the recognition scores of the index system among assistant physicians with different working years, educational background, professional title and work unit ( F/H=6.41, 14.83, 12.45, 7.53, P<0.01). Educational background(associate degree: B=0.091, P=0.015; bachelor degree and above: B=0.196, P<0.001) and professional title(intermediate professional title and above: B=-0.234, P<0.001) were the independent factors influencing the recognition degree of the index system for rural general practice assistant physicians. Conclusions:The post competency index system is generally recognized by rural general practice assistant physician, and academic qualifications and professional title status may influence its recognition.
Résumé
Purpose/Significance The paper discusses the application of artificial intelligence technology to the key entity recognition ofunstructured text data in the electronic medical records of lymphedema patients.Method/Process It expounds the solution of model fine-tuning training under the background of sample scarcity,a total of 594 patients admitted to the department of lymphatic surgery of Beijing Shijitan Hospital,Capital Medical University are selected as the research objects.The prediction layer of the GlobalPointer model is fine-tuned according to 15 key entity categories labeled by clinicians,nested and non-nested key entities are identified with its glob-al pointer.The accuracy of the experimental results and the feasibility of clinical application are analyzed.Result/Conclusion After fine-tuning,the average accuracy rate,recall rate and Macro_F1 ofthe model are 0.795,0.641 and 0.697,respectively,which lay a foundation for accurate mining of lymphedema EMR data.
Résumé
Objective To observe the value of a YOLOX target detection model for automatically identifying endovascular interventional instruments on images of digital subtract angiography(DSA).Methods DSA data of 37 patients who underwent abdominal endovascular interventional therapy were retrospectively analyzed.Totally 4 435 DSA images were captured and taken as data set,which were divided into training set(n=3 991)and verification set(n=444)at the ratio of 9∶1.Six kinds of endovascular interventional instruments were labeled.YOLOX algorithm was applied for deep learning of data in training set in order to build a target detection model,and the efficacy of the model for automatically identifying endovascular interventional instruments on DSA images was evaluated based on varification set.Results A total of 6 668 labels were put on 4 435 DSA images,aimed on Terumo 0.035in loach guide wire(n=587),Cook Lunderquist super hard guide wire(n=990),Optimed 5F with graduated pig tail catheter(n=1 680),Cordis MPA multi-functional catheter(n=667),Boston Scientific V-18 controllable guide wire(n=1 330)and Terumo 6F long sheath(n= 1 414),respectively.The training set contained 527,875,1 466,598,1 185 and 1 282,while the verification set contained 60,115,214,69,145 and 132 the above labels,respectively.The pixel accuracy of YOLOX target detection model for automatically identifying the above instruments in the verification set was 95.23%,97.32%,99.18%,98.97%,97.60%and 98.19%,respectively,with a mean pixel accuracy of 97.75%.Conclusion YOLOX target detection model could automatically identify endovascular interventional instruments on images of DSA.
Résumé
Objective:To explore the developmental characteristics of event-related potential(ERP) in cognitive function of recognition memory in children aged 6-12.Methods:A total of 130 normal children were divided into seven age groups (6 ( n=20), 7 ( n=17), 8 ( n=23), 9 ( n=24), 10 ( n=19), 11 ( n=15), and 12 years old ( n=12)) to perform a picture study-recognition task and record the reaction time, accuracy, and ERP components of all participants. SPSS 22.0 software was used for data analysis. Single factor analysis of variance and trend of variance were used to compare the response time and accuracy of 7 groups of children during the recognition stage. Pearson correlation analysis was used to study the correlation between the amplitude of the central midline N2 component and age. Paired t-test was used to examine the old/new effects of the amplitude of midfrontal N2 and midparietal P3 waves. Results:(1) The differences of recognition ability ( F(6, 123)=2.476, P<0.05), old picture reaction time ( F(6, 123)=6.461, P<0.001), and new picture reaction time ( F(6, 123)=4.163, P<0.001) among 7 age groups of children were statistically significant. Recognition ability of children aged 6 (0.61±0.24) was lower than those of 8-12 years old children((0.76±0.27), (0.76±0.10), (0.73±0.11), (0.75±0.10), (0.70±0.17) respectively)(all P<0.05). The reaction time of the old picture showed no difference among the children aged 6-9 (all P>0.05), and the reaction time of old picture of children aged 12 was shorter than those of 6-10 years old children (all P<0.01). There was no significant difference in the reaction time of new pictures among the children aged 6-10 (all P>0.05), and which in children aged 12 was shorter than those in 6-10 years old children(all P<0.01). (2) Age was positively correlated with the amplitude of the N2 component in the central region under the new ( r=0.488, P<0.001) and old picture( r=0.452, P<0.001) conditions. (3)At 6 years old, children showed old/new effects on the mid-frontal electrodes. At 7 years old, there were no old/new effects in either the mid-frontal or mid-parietal regions. From 8 to 9 years old, old/new effects appeared in the mid-parietal lobe. At 10 years old, old/new effects were present in both the mid-frontal and mid-parietal regions. At 11 years old, the mid-parietal lobe showed old/new effects. Finally, at 12 years old, negative old/new effects could be observed in both the mid-frontal and mid-parietal regions. Conclusion:There are three periods of changes in the behavior of picture recognition memory in school-age children. At ages 6-7, the accuracy rate is relatively low; at ages 8-9, it improves; and between ages 10-12, the accuracy rate stabilizes while also enabling faster judgments.Children's recognition memory retrieval process is more complex than their behavioral performance. Children have different tendencies toward strategies, but strategic transitions in recognition processing are not always beneficial for performance.
Résumé
Objective To establish a method for simultaneous determination of 11 components of Solanum nigrum from different producing areas,and to evaluate the quality by chemometrics and entropy weight-technique for order preference by similarity to ideal solution(EW-TOPSIS).Methods The 17 batches of Solanum nigrum samples from 8 provinces were collected.The high performance liquid chromatography(HPLC)method was used to simultaneously determine the contents of medioresino,pinoresinol,quercetin,rutoside,solasonine,solamargine,khasianine,solasodine,desgalactotigonin,diosgenin and β-sitosterol,and the multi-components quantitative control mode of Solanum nigrum was established.The quality evaluation model of Solanum nigrum was established by using chemical recognition pattern and EW-TOPSIS method,and the overall quality was evaluated comprehensively.Results When the 11 components were in the 0.78-39.00,0.55-27.50,0.34-17.00,0.21-10.50,41.87-2 093.50,60.95-3 047.50,2.58-129.00,1.02-51.00,0.46-23.00,1.05-52.50 and 0.42-21.00 μg/mL(r＞0.999 0),their linear relationships were good.The average recovery was 96.81％-100.28％with the RSD＜2.0％(n=9).17 batches of samples clustered into 3 categories.Solamargine,solasonine,desgalactotigonin and medioresino may be the main potential markers affecting the quality of Solanum nigrum.The results of EW-TOPSIS method showed that,the quality evaluation closeness of 17 batches of Solanum nigrum were 0.433 6,0.416 8,0.624 2,0.500 8,0.479 1,0.636 1,0.568 3,0.250 0,0.190 9,0.222 1,0.170 7,0.720 0,0.698 3,0.744 7,0.717 9,0.720 9 and 0.718 3,respectively,indicating that the overall quality of Solanum nigrum from Liaoning,Jilin and Heilongjiang were better,followed by Jiangsu,Henan and Anhui.Conclusion The established HPLC method for simultaneous determination of 11 components in Solanum nigrum is convenient and accurate.Chemometrics and EW-TOPSIS method are objective and comprehensive,which can be used for the overall quality evaluation of Solanum nigrum.
Résumé
Neuro-enhancement，by intervening on nerves for non-medical purposes，improves people’s physical，mental，and cognitive functions. While benefiting people，it also raises ethical risks of privacy，fairness，autonomy，and identity recognition between themselves and “artificial life”. Faced with these serious ethical risk challenges，it is urgent to propose countermeasures that respect and safeguard the basic rights of human beings，promote fair benefits with the principle of priority，standardize the information dissemination of neuro-enhancement，strengthen public education and training of ethical on neuro-enhancement technologies and advance responsible innovation，as well as carry out ethical education for neuroscience practitioners，with a view to promoting the healthy development of the field of neuroscience.
Résumé
With the development of more and more multi-center and cross-field cooperative medical research，the establishment of high-quality and efficient ethical collaborative review and mutual recognition systems is an inevitable demand for multi-institutional research，and an inevitable move to implement relevant national policies.Based on the work practice of ethical collaborative review and mutual recognition in Shenzhen，by analyzing the practical challenges of ethical collaborative review and mutual recognition in China，this paper proposed that to ensure the homogeneity and efficiency of review.Government departments need to take the lead，establish an ethical review alliance，and clarify responsibilities and rights.Based on actual needs，system first，and effective communication，ethical collaborative review and mutual recognition of results could be jointly promoted，aiming to provide a reference for our counterparts in China to promote ethical collaborative review and mutual recognition of cross-institutional research.
Résumé
Crossmodal transfer is the ability to apply the knowledge acquired in one sensory modality to another. Researches on crossmodal transfer investigate how the brain represents information from different sensory modalities, and provide new insights to improve cognitive processing efficiency and reduce repeated learning. To clarify the characteristics and mechanism of crossmodal transfer, this article first introduced the crossmodal transfer effect in different field of research, such as object recognition, category learning, and time perception. After that, the theoretical researches on the representation type of crossmodal transfer were reviewed, mainly including multisensory theory and multisensory mental imagery theory as well as the supportive and opposite findings. The research progresses on the neural mechanism of crossmodal transfer using ERP and fMRI techniques were introduced, mainly including metamodal theory, and multisensory reverse hierarchy theory as well as the supportive and opposite findings. The objective and subjective factors which influenced crossmodal transfer effect were sorted out, in which we suggested that the modality dominance phenomenon supports the metamodal theory, while other factors such as sensory experience, age, setting of learning tasks and stimulus features support theories such as the multisensory hypothesis. Finally, we described the potential applications of the current research findings on crossmodal transfer and pointed out future research questions in this field.
Résumé
The application of face recognition technology is gradually expanding to the medical field. It has been initially used in the medical diagnosis of endocrine diseases and genetic syndrome. This technology is expected to be used for the screening of genetic syndrome and endocrine diseases, shortening the delay period of disease diagnosis and helping the staging of endocrine diseases. However, this technology also has some moral risks, such as the risk of personal information security disclosure, the challenge to the future of mankind and the division of moral responsibility. This paper reflected on the dilemma of moral responsibility in the application of face recognition and medical diagnosis, and explored the two basic problems of "who is the subject of moral responsibility" and "the specific division of moral responsibility of different moral subjects" in face recognition and medical diagnosis. Finally, some suggestions on the moral responsibility in face recognition and medical diagnosis are put forward. The first is to determine the role of face recognition and medical diagnosis as an auxiliary category, and doctors are still the main medical subject; the second is to build the responsibility ethics mechanism and laws and regulations, the establishment of the responsibility system of face recognition and medical diagnosis is inseparable from the joint action of law and ethics.
Résumé
For multi-center clinical research, how to ensure the quality of ethical review and improve the efficiency of ethical review through cooperation among centers is an important direction for clinical research management departments and research parties to explore. By combing and analyzing the existing pattern of multi-center ethical review at home and abroad, combining the current situation of the ethical review and management development in China, taking cancer clinical research as the breakthrough point, it was advocated to establish a cooperative review led by professional institute in domestic, on the basis of extensive and in-depth training exchanges and effective communication on the same platform, collaborative review, ensure quality and efficiency, so as to promote and implement the "mutual recognition" of ethical review. Then, this paper further put forward the concept of "whole-process linkage" in the ethical management process of multi-center clinical research, and pointed out that all research parties should clarify their responsibilities, enhance their awareness and ability, and jointly and comprehensively implement the protection of subjects among clinical researchers.