MCU和MEMS将用于预测预防老年人跌倒

对年长者来说，若跌倒并导致受伤可不是开玩笑的；为此美国德州理工大学(Texas Tech University)在德州仪器(TI)的赞助下展开了一项研究，目标是藉由分析年长者的姿势与步伐，好在他们可能跌倒之前发出警告。 这项研究项目已经在德州理工大学的健康科学中心(Health Science Center)征求自愿者，以进行穿戴式无线感测装置的最佳化，接下来将在美国首个大学附设的老人病学研究机构──也就是德州理工大学老人病学教育与看护中心(Geriatric Education and Care Center)──进行临床实验。 “我们已经尝试过各种方式的感测装置布置法，例如穿戴在腿部的惯性传感器，以及鞋底内含压力传感器的拖鞋；而德州仪器的低功耗微控制器与结合无线电的人体穿戴 MEMS 加速度计/陀螺仪──就像外接式的心律调整器──效果最好。”该研究项目的主持人、德州仪器电子电机教授Donald Lie表示。 经过三年的开发时间，Lie的研究团队不只打造出人体穿戴式无线感测装置，也研发了一套能无线监测病患的PC软件，所产生的算法能可靠侦测患者跌倒的方向。Lie表示，要分辨出患者是从床上、汽车座椅或是家中各处的家具上跌落，是很困难的。

M7gesmc

可侦测年长病患跌倒的无线感测装置模块原型，其小尺寸可以夹在皮带上 接下来，德州理工大学的研究团队将藉由在该校老人病学教育与看护中心的病患身上进行临床实验，将其软件分析效果更精细化，好侦测出跌倒之前的征兆，达到预防的效果。新开发出的无线MEMS感测装置内含加速度计、陀螺仪，可以夹在皮带上，若是不配戴皮带的患者或是女性，也可以夹在内衣背后。 该装置采用了TI的MSP430微控制器、CC2500射频(RF)收发器，支持低功耗SimpliciTI网络通讯协议，执行以TI ez430-RF2500开发平台开发的专属软件堆栈。TI院士Allen Bowling表示，此研究成果受到瞩目的原因，是不只能侦测病患是否跌倒，还可能提供预防性警告：“藉由分析病患在跌倒之前的姿势与步伐，我们希望能在病患跌倒之前发出警告，好让他们能及时抓住东西或是坐下。” 而若无法阻止患者跌倒，该装置也能立即发送无线讯号给照顾者；但该研究项目的目标是将分析方法──目前是在医疗机构负责监控的PC上执行──编码入MSP430微控制器中，因此无论病患是否位于可接收无线讯号的范围中，都能收到警告讯息。 此外研究团队也将针对其它与人体平衡相关的疾病患者进行监测，例如帕金森氏症、失智症以及癫痫等，期望也能提供这些患者预防性的警告讯息。“我们相信这些研究项目能为老年人的临床看护带来很大进展。”Lie表示；其研究团队成员还包括Tam Nguyen、Steven Zupancic、 Andrew Dentino、Ron Banister与Tim Dallas等多位医师。 编译：Judith Cheng 本文授权编译自EE Times，版权所有，谢绝转载 参考英文原文：MEMS project aims to prevent elderly from falling，by R. Colin Johnson

相关阅读：
• Q1 20大半导体厂商排名，高通超越TI
• 几大MEMS制造商排名可能很快发生剧烈变化
• 更快更高更强！MEMS在运动领域M7gesmc

{pagination} MEMS project aims to prevent elderly from falling R. Colin Johnson PORTLAND, Ore.—The infamous 1980s television commercial that featured the tag line "I've fallen and I can't get up" became the butt of a thousand jokes. But for the elderly, susceptibility to falls and resulting injuries is no laughing matter. Now, a development effort at Texas Tech University, sponsored by Texas Instruments Inc., is taking aim at preventing falls by analyzing posture and gait to send warning alerts to the elderly before they fall. The project has already enlisted volunteers at Texas Tech's Health Science Center to perfect the wireless wearable sensor and is on track next for clinical trials at the first U.S. on-campus geriatric teaching facility, Texas Tech's Geriatric Education and Care Center. "We have tried all sorts of sensor placements, from leg-mounted inertial sensors to slippers with pressure sensors in their soles, but Texas Instruments' low-power microcontrollers and wireless radio combined with a torso-mounted MEMS accelerometer and gyro—like an external pacemaker—gives the best results," said lead scientist on the project, Texas Tech EE professor Donald Lie. After a three-year development effort, Lie's team has crafted not only the torso-mounted wireless sensor but also the software analytics running on a PC that wirelessly monitors patients, resulting in algorithms that can reliably detect falls regardless of in which direction, which Lie claims is difficult to differentiate from the many ways that people plop down into their beds, cars seats and the various pieces of furniture around the home. Next, the team is honing its software analytics even finer to detect pre-fall conditions in order to take preventative measures by virtue of clinical trials on real patients at the local Geriatric Education and Care Center where they have already installed the necessary wireless infrastructure. The MEMS sensor and wireless radio module is clipped on the belt, or for women not wearing a belt on the back of their bra, and contains a MEMS accelerometer, MEMS gyroscope, TI's MSP430 microcontroller and CC2500 radio frequency (RF) transceiver and uses the ultra-low-power SimpliciTI network protocol running a proprietary software stack developed on TI's ez430-RF2500 development platform. "We became interested in this project at Texas Tech, because it went beyond just fall detection, but aspired to preventative measures," said TI Fellow Allen Bowling. "By analyzing the dynamics of posture and gait--the way people are standing or walking before a fall—we hope to be able to send an alarm instructing them to grab hold of something or sit down before they fall." If the patient does end up falling, an alarm is sent wirelessly to a care provider, but the goal is to code the analytics—which today runs on a clinician-monitored PC—for the MSP430 microcontroller so alarms can be issued to patients regardless of whether they are in range of a wireless router or not. The team also has its sights on monitoring other vestibular (balance related) disorders and diseases including Parkinson's, dementia and epilepsy, in hopes of issuing warnings to those patients of impending episodes thereby enabling them to take preventative measures. "We believe that these projects will really make a significant difference in the clinical care of geriatric populations," said Lie, who credits his team of collaborators, including doctors Tam Nguyen, Steven Zupancic, Andrew Dentino, Ron Banister, and Tim Dallas.