实施健康中国行动,提高全民健康水平

7月15日,国务院正式发布

《关于实施健康中国行动的意见》,

这是国家层面指导未来十余年

疾病预防和健康促进的一个重要文件。

依据《意见》

成立健康中国行动推进委员会

印发《健康中国行动(2019—2030年)》

《健康中国行动组织实施和考核方案》

聚焦当前人民群众面临的

主要健康问题和影响因素

从政府、社会、个人(家庭)

3个层面协同推进

通过普及健康知识、参与健康行动、

提供健康服务,实现促进全民健康的目标

从以“疾病”为中心

向以“健康”为中心转变

从注重“治已病”

向注重“治未病”转变

从依靠卫生健康系统

向社会整体联动转变

从文件向社会倡议转变

实现政府牵头负责

社会积极参与

个人体现健康责任

健康中国“共建共享”

“把健康融入所有政策” 

@所有医生:医师节快乐!

  • 2019年8月18日,中国医师节庆祝大会在人民大会堂举行。在庆祝大会上,由中国工程院院士乔杰领誓,会场千余名医生集体宣誓:我志愿献身人类的健康事业;自觉维护医学的尊严和神圣;敬佑生命,救死扶伤,平等仁爱,尊师重道;诚实守信,恪守医德,精益求精,慎思笃行;以上誓言,源于心,践于行!
  • 2018年8月18日,习近平对首个“中国医师节”作出重要指示强调:弘扬救死扶伤的人道主义精神,不断为增进人民健康作出新贡献
  • 2017年8月18日,国务院通过了卫计委关于“设立中国医师节”的申请,同意自2018年起,将每年的8月19日设立为“中国医师节”。中国医师节是经国务院同意设立的卫生与健康工作者的节日,体现了党和国家对1100多万卫生与健康工作者的关怀和肯定。

广东省基层医药学会人机大战圆满结束,我司取得优异成绩!

2019年8月4日,经过三个多月紧锣密鼓地筹备工作,“面向基层,广接地气”的广东省基层医药学会细胞病理与分子诊断专业委员会成立大会暨学术会议顺利召开并圆满结束!来自广东省各大医院的细胞病理与分子诊断学科领域的专家、教授、学者,以及一线的专业人员约280人汇聚一堂,共同见证了专委会的成立以及备受瞩目的宫颈癌细胞病理筛查“人机大战”的圆满举办,我司参加了人机大赛并取得了优异的成绩。

人机大战由人工智能厂商对战基层病理筛查员,大赛裁判组由业界知名细胞病理专家:欧阳能太教授、余俐教授、曹箭教授组成,比赛规则如下:

(主持人介绍比赛规则)

1.比较人工筛片与机器筛片的准确性,不比较快慢,只看正确与否,与标准答案一致为正确,得1分,不一致判断为错误,得0分。

2.参与者包括筛查员、AI企业以及观战团(观战团观众持答题器)。

3.筛查员分5组与AI进行PK,3人一组,共15张液基细胞玻片,每人一张。每组阅片时间3分钟。

经过5轮的比赛过后,我司以较大的优势胜出!

(由于比赛时间较短,15张病理玻片数量较少,不能充分展现筛查员和人工智能的真实技术水平,但从比赛流畅度、契合度,比赛气氛等多个方面来讲,本次大赛非常成功,我司表现异常优秀,从比赛开始后无任何卡顿,我司辅助诊断结果准确。大赛结果可喜,在一定程度上验证了人工智能在基层筛查等应用场景和能力)
赛后,学会为参赛筛查员和参赛企业颁发了奖牌,全体参赛员与裁判组合影

赛后专家点评:欧阳能太教授提到通过短时间筛查评判片子是有难度的,这是人机大战第一次亮相,通过15例能让大家见识到AI技术的发展,旗鼓相当,结果可喜;邵建永教授则表示人工智能是未来是病理医生的好帮手,人工智能深度发展势不可挡;王连唐教授说道AI技术不断进步,但是速度不重要,关键是要准确,AI未来不能代替人,但AI对简单的病理筛查是有能力胜任的,基层医院要重视。

Digital cytopathology: a constant evolution

Simone L. Van EsVanessa WhiteJennifer RossJanelle GreavesStephanie GayDerek HolzhauserTony Badrick 

转自 Digital cytopathology , https://doi.org/10.1111/cyt.12554

Abstract: Capitanio et al provide a succinct review of digital cytopathology technology with particularly good overviews of scanner focusing methodologies as well as explanations of deep focusing techniques. The aim of this letter is to briefly comment and expand on some of the information provided by these authors.

The field of digital cytopathology is in constant evolution, the main focus being how to accommodate the z-axis cell distribution. Capitanio et al provide a succinct review of digital cytopathology technology with particularly good overviews of scanner focusing methodologies as well as explanations of deep focusing techniques. The aim of this letter is to briefly comment and expand on some of the information provided by these authors.

In an ideal world spacing between z-stacks should be tailored to each individual specimen to account for the unpredictable and varied distribution of cells in the z-axis. However, this is impractical and clear evidence-based guidelines for appropriate z-axis parameters are delayed. From a practical perspective, topography software certainly assists in the modification and selection of such scanning parameters. The Royal College of Pathologists of Australasia Quality Assurance Programs (RCPAQAP) is increasingly z-stacking their cytopathology slides in entirety2 , utilising Sectra Uniview to view these z-stacked images through smart streaming. Trials are underway to determine the efficiency of using this z-axis technology for screening and reporting cytopathology.

Capitanio et al discuss optimizing whole slide image quality through manipulation of software. However the challenges of cytopathology WSI (whole slide imaging) can certainly be partially circumvented by modifying cytology specimens themselves so that they are better suited to scanning and image acquisition. Simple solutions such as rapid on-site evaluation of cytology specimens to remove tissue fragments for embedding or use of cell block preparation which presents a more suitable cell distribution topography, is a straightforward way to improve the quality of final WSI.

Focus-fusion or extended focus imaging (EFI) discussed by Capitanio et al – where multiple z-axis planes are assimilated into a single plane – is tempting technology.1 The positive is that the final file for the EFI uses the memory and transfer requirements of a single plane scan. Negatives include the scanning time (which is still the same as that required for creating a multi-layered z-stack) as well as data loss (once all information is combined into EFI, data from the separate z-axis planes is discarded permanently). There can also be problems with background and resolution. Cells that were blurry in the multilayer stack are brought into focus with EFI, but the EFI algorithm is also applied to the extracellular background debris throughout the thickness of the slide potentially resulting in a grainier appearance. This is why sematic focus point analysis, as described in the review by Capitanio et al, may hold promise. Here software integrates cells in a 3-dimensional capacity that are sharply in focus, avoiding artefact and out-of-focus objects. Capitanio et al then takes this thought-process one step further suggesting compression methods that “discard the redundant material that is replicated between z-stack layers, saving only the differences between them”.

Capitanio et al are correct in stating that there is a lot of negative perception concerning digital cytopathology. However, digital pathology, especially digital cytopathology, has the potential to reverse the “absence” of pathologist input in medical education as well as to improve the quality and equity of medical educational resource availability. Capitanio et al cite the numerous advantages afforded by digital cytopathology for medical education. Additionally, WSI can be easily incorporated into effective and efficient E-learning platforms. Interactivity when teaching pathology, a feature easily promoted through use of WSI, has been shown to improve engagement, performance in examinations, as well as increase cytopathology diagnostic accuracy skills compared to learning with glass slides and a microscope. Capitanio et al tabulate a helpful range of digital cytopathology educational resources that are available online. More recently, The Royal College of Pathologists of Australasia have also developed a Cytopathology eCase library containing hundreds of cytopathology WSI. Many of the slides have been scanned entirely in the z-axis and can be viewed seamlessly in the z-axis with Sectra Uniview. Access to this educational library is restricted to RCPA trainees and pathologists.

In response to discussions on proficiency testing raised by Capitanio et al, we would also like to respond from the Australian perspective. The Royal College of Pathologists of Australasia Quality Assurance Programs (RCPAQAP) has been offering proficiency programs globally in gynaecological and non-gynaecological cytopathology for more than 20 years, with WSI cases being introduced as a mandatory part of the non-gynaecological cytopathology modules since 2011.

Negative perception towards digital cytopathology can hopefully be partially addressed by advancing z-axis technology. Complex computational filtering of cellular material from multiple focal planes with incorporation into WSI with a single focal plane, may hold promise. We certainly still need a convenient way of annotating z-stacked WSI. Improvements in viewing technology as well as workstation ergonomics will address some of the current barriers to efficient screening of cytopathology WSI. Modifying pre- and post-scanning workflow may also be helpful. Such improvement are already making a difference to education and proficiency testing.

卫健委:3年内用AI等手段将”两癌”筛查覆盖率提升27%

经济观察网 记者 瞿依贤 7月25日,健康中国行动推进委员会办公室召开新闻发布会。记者在会上获悉,卫健委欲将人工智能技术引进“两癌”筛查,且3年内将“两癌”筛查覆盖率提升27%。

  健康中国行动(2019-2030年)之妇幼健康促进行动有多项目标,其中一项是:农村适龄妇女宫颈癌和乳腺癌(下称“两癌”)筛查覆盖率到2022年和2030年分别达到80%及以上和90%及以上(覆盖率以县为单位统计)。以基线值52.6%计算,到2022年筛查覆盖率至少提高27%,到2030年覆盖率至少提高37%。

  宫颈癌和乳腺癌一直是女性高发癌症。2019年1月,国家癌症中心发布了最新一期的全国癌症统计数据,恶性肿瘤高发疾病中,女性发病首位为乳腺癌,第6位为宫颈癌。

  从促进行动的结果性指标来看,当前“两癌”防治与目标水平之间存在较大差异。对此,国家卫生健康委妇幼司司长秦耕表示:“国际社会逐步关注宫颈癌防治,推动消除宫颈癌,我们国家‘两癌’发生率还是在逐步上升,形势不容乐观。”

秦耕指出,要从三方面来推动“两癌”筛查:加强健康教育,提高自我保护意识,组织和动员适龄妇女及时自觉接受政府提供的筛查服务;建立多元立体“两癌”防治体系,妇幼保健机构、综合医院、技术指导单位、妇联和社会组织多方合作,搭建综合防治体系;以农村为重点,逐步扩大“两癌”检查覆盖面,推动地方政府来给予更多投入。

  发布会上还公布了一组数据:2009年,原国家卫生部、财政部、全国妇联发布《农村妇女“两癌”检查项目管理方案》,在全国范围内(221个县区)对农村地区35-64岁的适龄妇女进行“两癌”检查,至今宫颈癌检查达近1亿人次,乳腺癌的检查超过了3千万人次,检出宫颈癌及癌前病变17.7万例,乳腺癌和癌前病变1.6万例。此外,目前宫颈癌检查已覆盖全国2118个县,乳腺癌检查已覆盖全国1651个县。

  毛群安表示,农村地区妇幼健康服务的力量本身较紧张,基层承担着很大压力,为了提升肿瘤筛查的准确性,减轻基层负担,卫健委正在支持相关医学人工智能技术的开发,“希望基于过去的检查、大数据,开发出适合我国的人工智能手段,这样就能够减轻基层的工作压力,便于更大规模地开展癌症筛查。”

源智医生,辅于病理-塞维森参加2019年8月4号:广东省基层医药学会细胞病理与分子诊断专委会成立大会暨2019逸仙细胞分子学术论坛人机大赛

近几年,人工智能在各个领域崭露头角,前有阿尔法Go大败人类棋手,后有各类手术机器人大显身手。科技正在颠覆我们的想象, AI+医疗已经步入更加深入的融合。妇科宫颈细胞筛查对于降低女性宫颈癌发病率至关重要,由于癌前病变细胞检测对病理专家的经验和水平要求较高,加上病理医生的稀缺,世界各国都希望通过人工智能等技术协助医疗专家进行筛查工作。

于2019年8月4号,广东省基层医药学会细胞病理与分子诊断专委会成立大会期间,我司受邀参加大会《人机大赛》环节,将与人工智能友商和病理筛查医生同台竞技,接受广大病理医生的考验和指导。

干杯赛维森-庆51小团建

赛维森(广州)医疗科技服务有限公司致力于通过人工智能辅助细胞病理诊断,秉承聚焦细胞、专注病例、医工结合,砥砺前行之宗旨,通过技术服务健康事业。公司于2019年5.1放假前夕组织团队建设,以酒壮医胆,鼓励团队前行!


赛维森(广州)医疗科技服务有限公司成立

赛维森(广州)医疗科技服务有限公司(以下简称:赛维森医疗科技)于2019年1月30日成立,是一家将人工智能应用于细胞病理领域的创新型公司,核心是依托人工智能深度学习、数字病理与医学影像分析技术,对病理大数据进行智能化识别和分析,快速、精确地提供辅助诊疗方案建议,从而提高临床诊疗的精准度和效率,可广泛应用于医院、独立影像和检验中心、远程诊断云平台、医学成像设备厂商、医学教研机构和患者等用户群体.

联系我们


联络地址

2313, 达镖国际,
江南大道中362号
海珠区 广州市

请留下您宝贵意见建议或投诉