Skill-style-support analytics
Skill: knowledge of operation and best practices of healthcare, life science, bio-technology, pharmacy, technical, system administration, management, governance, supply chain management;
Style: leadership, shared vision, goal setting, intelligent communication, risk assessment and mitigation, innovation project management;
Support : proactive, preventive and reactive support.
The expert panel are analyzing skill-style-support for health security. What should be the innovation model for effective diffusion of Cancare ? is it possible to adopt K- A-B-C-D-E-T-F model? The workforce involved in aforesaid technological innovations are expected to develop different types of skills in technical (e.g. bio- medical engineering, pharmacy, life-science, oncology, deep learning and artificial neural network), healthcare and medical science domain such as research and development, knowledge management, product design, project management, supply chain management, sales and distribution. It is essential to teach deep learning innovatively in various programmes of Electrical, Electronics and Biomedical engineering as part of graduation, post graduation and Doctoral programmes. The learning community should be involved in consulting, projects and research assignments. They need good resources such as digital libraries having good collection of books, journals and magazines, software and experimental set up. The workforce of R&D labs can develop skills through effective knowledge management programmes and resources which support creation, storage, sharing and application of knowledge. The diffusion of technology requires the support of intelligent leadership style; the leaders must be able to tackle the complexity, pace and novelty of R&D projects through efficient project management, organization structure development, knowledge management and collaborative and cooperative work culture. The healthcare professionals are expected to be people, information and action oriented. Next, let us consider the element support.
Caution from malicious learning system : The basic objective is to protect learning systems in adversarial setting from various types of threats such as use of flawed learning algorithm or intentional change of training and testing data distribution. The malicious agents may act consciously to limit or prevent accurate performance of the learning system for economic incentives. It is a common problem where machine learning is used to prevent illegal or unsanctioned activities. Traditional techniques (e.g. efficient algorithm, linear classification) are necessary but not sufficient to ensure the security of the machine learning system. It is a hard problem and needs the support of an efficient mechanism equipped with intelligent threat analytics and adaptive secure multi-party computation algorithms. Malicious business intelligence is a critical threat to machine learning system. The conflict between security intelligence and business intelligence is inevitable. It needs fair, rational and intelligent business model innovation.
Example : Malicious business intelligence may attack a life-science supply chain and healthcare service chain through greedy heuristics in payment function for revenue and profit optimization, economic pressure and incentive policy, fraudulent health insurance model, flaws in investment decision on technology management, irrational and dull HR policy in talent management and chaotics in formulation of public policy, mechanisms and corporate governance. In fact, the conflict between business intelligence and security intelligence is inevitable; the deep learning mechanism is expected to resolve this conflict between security and business intelligence.
Let us consider a specific instance of machine learning in healthcare service chain. The deep learning mechanism must call the threat analytics to audit various critical processes associated with a healthcare service chain in cancer care such as registration, consulting, testing, surgical operations, billing, payment processing and follow-up. Generally, different types of information systems are commonly used to support these processes such as transaction processing system (TPS), decision support system (DSS), group decision support system (GDSS), knowledge management system (KMS) and business intelligence (BI) system. The primary objective of these information systems is to ensure fairness and correctness in computation of registration card, appointment slip for consulting, prescription by consultant, surgery schedule, quality control certificate, medical test report, discharge certificate, bills and payment receipt, feedback form and patient’s guide. The other important issue is to preserve the privacy of patient’s personal and medical data. The deep learning mechanism should verify the security of the computing schema associated with the machine learning system in healthcare service chain to identify probable sources of errors in cancer care.
Incorrect data provided by the cancer patients to the registration associate during registration intentionally or due to lack of knowledge or incorrect perception of the patients or their attendants; the patients or their attendants may be irrational in information sharing properly with the service providers.
No verification of patient’s identity correctly during registration; the cases of emergency situation or accidents may skip verification due to unavailability of data about the patients.
Wrong entry of data into various information systems by the healthcare associates due to time and resource constraints or misunderstanding or lack of validation of input data.
Computational errors due to wrong configuration of enterprise applications and
/ or errors in the heuristics, deep learning algorithms and quantitative models and / or no updating of data (e.g. service charge, tariff of testing, price of drugs and healthcare products; low accuracy of pattern recognition algorithms in image processing system may result incorrect medical diagnosis.
Access control problem causing dangerous errors in information system; a malicious agent may enter false data into HIS during the absence of authorized users.
A malicious agent may launch attacks on TPS, DSS, GDSS, KMS and BIS through malicious data mining, insecure data storage, flaws in data visualization and image processing algorithms and transaction processing logic.
Swap or mixing of test data of various patients or drugs administration due to confusion, poor document management, lack of clear understanding or training of the healthcare workforce; false data injection on viruses in test reports are serious threats in today’s healthcare practice. The patients are not often given test reports today by the service provider to hide malicious trading practice or to charge extra amount. Testing of uncommon viruses enhance the cost of testing unnecessarily. Sometimes, broadcast of epidemic results panic among the public and this critical and helpless situation is exploited by malicious testing and medicare practice inhumanly.
Errors in decision making by the health consultants due to lack of proper knowledge management system (e.g. case based reasoning, intelligent DSS and GDSS) or misperception or lack of coordination among the workforce of various departments or inappropriate enterprise application integration or error in test reports; incomplete prescription due to memory failure or silly mistakes.
Errors in scheduling due to exceptions (e.g. unfit patients, non-availability of healthcare experts), flawed and inadequate doctor-patient ratio;.
surgical operation by unauthorized and unskilled workforce, intentional errors due to malicious business practice, lack of ethics, casual approach and dull HR policy; unintentional errors due to physical and mental fatigue for excessive workload and sickness, non-availability of basic infrastructure and logistics arrangements;
Lack of verification of correctness of computation in medical billing and payment processing by the service provider and / or service consumer;
Incorrect data in patient’s help guide may cause confusions and mismatch between the computed results and perceived one;
Incorrect feedback by the patients or their attendants due to misperception, misunderstanding of feedback form, lack of knowledge and critical observations or casual attitude.
Sybil attack: It is really complex to trace the corrupted players in healthcare domain. A malicious agent may control multiple pseudonymous identities and can manipulate, disrupt or corrupt an application that relies on redundancy by injecting false data or suppressing critical data; it is sybil attack. The patients may be treated incorrectly and diagnosed as cancer casually though there is another simple medical problem. Natural intuition and perception may not be applied for simple medical problems. The patients may be incorrectly recommended for costly treatment. They may be recommended for costly treatment procedure repeatedly (e.g. CT scan, X-ray), drugs and surgicaloperations. The poor and helpless patients may be forced to validate and verify the test reports and medical diagnosis at various healthcare institutes. This is an instance of modern biological, chemical and radiological terrorism today.
Fairness and correctness of computation and testing is a critical concern in cancer therapy. Knowledge management is another critical success factor; case based reasoning may be a good solution for correct clinical decision making. For effective deep learning system, digital technology management is not only the critical success factor (CSF). There are other several CSFs such as HR policy in talent management, motivation and commitment, quality of education in terms of trust, ethics and values, intelligent public policy, mechanisms and corporate governance.
The workforce involved in aforesaid technological innovations are expected to develop different types of skills in technical (e.g. bio-medical engineering, pharmacy, life-science), healthcare and medical science domain such as research and development, knowledge management, product design, project management, supply chain management, sales and distribution. It is essential to teach Biomedical technology innovatively in various programmes of Electrical, Electronics and Biomedical engineering as part of graduation, post graduation and Doctoral programmes. The learning community should be involved in consulting, projects and research assignments. They need good resources such as books, journals, software and experimental set up. However, they should understand the motivation of the problems and various issues of technology management through deep analytics. The workforce can develop skills through effective knowledge management programmes and resources which support creation, storage, sharing and application of knowledge. The diffusion of technology requires the support of intelligent leadership style; the leaders must be able to tackle the complexity, pace and novelty of R&D projects through efficient project management, organization structure development, knowledge management and collaborative and cooperative work culture. The leaders are expected to be people, information and action oriented.
It is essential to develop skill in new product development through proper coordination among design, supply and patient chain and R&D, production and marketing functions. The basic objectives are to maximize fit with customer’s needs and demands, ensure quality assurance, minimize time to market, and control product development cost. It may be an intelligent initiative to involve the suppliers and the customers in the development process, beta testing, fault tree and failure mode effects analysis and TFPG as part of quality control measures. It is really challenging to manage new product development team through diversity, knowledge base, multiple skills, problem solving capability, cooperative corporate culture and intelligent communication protocol at optimal administrative costs.
Let us analyze skill-style-support. The workforces involved in this technological innovation are expected to develop different types of skills in medical science, immunology, disaster operation management and system administration, innovation on life-science, pharmacy and biotechnology, research and development, testing, tracing, tracking, benchmarked and standardized medical practice. The orkforce can develop skills through effective knowledge management programmes. An effective knowledge management system supports creation, storage, sharing and application of knowledge in a transparent, collaborative and innovative way. The diffusion of the innovation requires the support of great leadership style, effective governance, formation of special task force and expert committee, intelligent and rational corporate communication. The style is basically the quality of leadership; the great leaders must have passion, motivation and commitment. The leaders must be able to share a rational vision, mission and values related to the innovation among all the stakeholders honestly and appropriately in time.
What should be the ideal organization model for this technological innovation? A traditional functionally centered organization model may not be suitable for supporting end-to-end healthcare process. Such process management is more than a way to improve the performance of individual processes; it is a way to operate and manage a business. An enterprise that has institutionalized process management and aligned management systems to support is a process enterprise. It is centered on its customers, managed around its processes and is aligned around a common, customer oriented goal. The business models of top technological innovations require the support of a process enterprise structure enabled with advanced information and communication technology. The structure should have project, design, production, supply chain management maintenance, human resource management, sales & marketing and finance cells. The structure should be governed by an executive committee comprising of CEO and directors. The process managers should be able to identify core processes in the value chain; communicate throughout the organization about these critical processes; create and deploy measures regarding end-to-end process performance and define process owners with end-to-end authority for process design, resource procurement, process monitoring for redesign and improvement. The process enterprise requires a collaborative and cooperative work culture. Innovation in edidemic and pandemic control demands proper technological support in testing and medical diagnosis, proactive, reactive and preventive support for proper technology management. The technology needs the support of a collaborative enterprise model.